From nscott@aconex.com Thu Nov 13 20:48:50 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=-1.8 required=5.0 tests=AWL,BAYES_00 autolearn=ham version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAE4mnRo021248 for ; Thu, 13 Nov 2008 20:48:50 -0800 X-ASG-Debug-ID: 1226638127-1e18012a0000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from postoffice2.aconex.com (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 14D4415688C6 for ; Thu, 13 Nov 2008 20:48:47 -0800 (PST) Received: from postoffice2.aconex.com (prod.aconex.com [203.89.192.138]) by cuda.sgi.com with ESMTP id CACy9GJwefmsPLst for ; Thu, 13 Nov 2008 20:48:47 -0800 (PST) Received: from postoffice.aconex.com (localhost [127.0.0.1]) by postoffice2.aconex.com (Spam Firewall) with ESMTP id 321E637AB50 for ; Fri, 14 Nov 2008 14:32:08 +1100 (EST) Received: from postoffice.aconex.com (postoffice.yarra.acx [192.168.102.1]) by postoffice2.aconex.com with ESMTP id 9kDlCHuG83nDiKQC for ; Fri, 14 Nov 2008 14:32:08 +1100 (EST) Received: from [192.168.5.24] (melho0.aconex.com [203.89.192.141]) by postoffice.aconex.com (Postfix) with ESMTP id 4280A92C312 for ; Fri, 14 Nov 2008 14:22:31 +1100 (EST) X-ASG-Orig-Subj: pcp updates Subject: pcp updates From: Nathan Scott To: pcp@oss.sgi.com Content-Type: text/plain Date: Fri, 14 Nov 2008 14:30:01 +1100 Message-Id: <1226633401.4355.37.camel@verge.scott.net.au> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: prod.aconex.com[203.89.192.138] X-Barracuda-Start-Time: 1226638129 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -2.02 X-Barracuda-Spam-Status: No, SCORE=-2.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests= X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.10203 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- X-Virus-Scanned: ClamAV version 0.91.2, clamav-milter version 0.91.2 on oss.sgi.com X-Virus-Status: Clean Changes committed to git://oss.sgi.com/nathans/pcp.git configure.in | 2 +- debian/changelog | 1 + src/include/platform_defs.h.in | 1 + src/libpcp/src/config.c | 33 +++++++++++++++++++++++++++++---- src/libpcp/src/win32.c | 12 ++++++++++++ src/pmie/rc_pmie | 2 +- src/pmlogctl/pmlogger_check.sh | 17 ++--------------- 7 files changed, 47 insertions(+), 21 deletions(-) commit c09642302379f4408417d9731d00a76a5bcdca73 Author: Nathan Scott Date: Fri Nov 14 14:02:31 2008 +1100 Fix accidental omission in rc_pmie wrt IS_ON env var - we no longer set this, so don't use it. Also cleanup a similar issue in the pmlogger cron scripts to do with IS_ON for IRIX platforms - can now be done consistently for all platforms (note also -h and -L equate, in both Solaris and Linux versions of the test(1) utility), whereas previously we'd have defaulted to back-compat IRIX behaviour (bad). commit fb21f2f14aef137a0c4608b45fe9aeb5b882aead Author: Nathan Scott Date: Fri Nov 14 13:56:34 2008 +1100 Make note in debian packaging about already-fixed gcc 4.4 build issues. commit 960674affd6d80cb913c2ca5bcdaffee49374378 Author: Nathan Scott Date: Fri Nov 14 13:55:54 2008 +1100 Implement rindex(3) on Win32 and use it for prefixing PCP_DIR onto standard PCP paths. commit 9e969b85aa5cdbd000d8544805fbed3ecaa5b006 Author: Nathan Scott Date: Fri Nov 14 13:53:01 2008 +1100 Use minialistic pcp_var_dir on win32, for installation simplicity. From nscott@aconex.com Fri Nov 14 21:38:52 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=-1.8 required=5.0 tests=AWL,BAYES_00 autolearn=ham version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAF5cqSd025966 for ; Fri, 14 Nov 2008 21:38:52 -0800 X-ASG-Debug-ID: 1226727527-0c97026c0000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from postoffice2.aconex.com (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 231DC10BFAFD for ; Fri, 14 Nov 2008 21:38:47 -0800 (PST) Received: from postoffice2.aconex.com (prod.aconex.com [203.89.192.138]) by cuda.sgi.com with ESMTP id KpWHPCXlGFKJcQcR for ; Fri, 14 Nov 2008 21:38:47 -0800 (PST) Received: from postoffice.aconex.com (localhost [127.0.0.1]) by postoffice2.aconex.com (Spam Firewall) with ESMTP id 0D5B5379E3A for ; Fri, 14 Nov 2008 14:31:15 +1100 (EST) Received: from postoffice.aconex.com (postoffice.yarra.acx [192.168.102.1]) by postoffice2.aconex.com with ESMTP id bBlA3X8dwbajOmhq for ; Fri, 14 Nov 2008 14:31:15 +1100 (EST) Received: from [192.168.5.24] (melho0.aconex.com [203.89.192.141]) by postoffice.aconex.com (Postfix) with ESMTP id 1BD2F92C2E7 for ; Fri, 14 Nov 2008 14:21:38 +1100 (EST) X-ASG-Orig-Subj: kmchart updates Subject: kmchart updates From: Nathan Scott To: pcp@oss.sgi.com Content-Type: text/plain Date: Fri, 14 Nov 2008 14:29:08 +1100 Message-Id: <1226633348.4355.34.camel@verge.scott.net.au> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: prod.aconex.com[203.89.192.138] X-Barracuda-Start-Time: 1226727532 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -1.52 X-Barracuda-Spam-Status: No, SCORE=-1.52 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests=BSF_RULE7568M X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.10284 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- 0.50 BSF_RULE7568M Custom Rule 7568M X-Virus-Scanned: ClamAV version 0.91.2, clamav-milter version 0.91.2 on oss.sgi.com X-Virus-Status: Clean Changes committed to git://oss.sgi.com/nathans/kmchart.git src/chart/fileformat.cpp | 1 + src/libqwt3d/qwt3d_function.cpp | 1 + src/query/kmquery.h | 2 ++ 3 files changed, 4 insertions(+) commit a6ab4886ed80191d2e08b181abfbb2f5aeb0f811 Author: Nathan Scott Date: Fri Nov 14 14:27:15 2008 +1100 Fix scale-setting from a config file, thanks to Nigel Donaldson. commit 471a36529a562357125f7af756ce0418bdde007f Author: Nathan Scott Date: Sat Nov 8 14:57:31 2008 +1100 Fix compilation under gcc 4.4, thanks to Martin Michlmayr. From nscott@aconex.com Mon Nov 17 17:38:55 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=-1.9 required=5.0 tests=AWL,BAYES_00 autolearn=ham version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAI1cswo010680 for ; Mon, 17 Nov 2008 17:38:54 -0800 X-ASG-Debug-ID: 1226972328-437d039e0000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from postoffice2.aconex.com (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id CE4731591FF8 for ; Mon, 17 Nov 2008 17:38:48 -0800 (PST) Received: from postoffice2.aconex.com (mail.aconex.com [203.89.202.182]) by cuda.sgi.com with ESMTP id DtaZxuTRzUQ4lYjg for ; Mon, 17 Nov 2008 17:38:48 -0800 (PST) Received: from postoffice.aconex.com (localhost [127.0.0.1]) by postoffice2.aconex.com (Spam Firewall) with ESMTP id 175D63A7303 for ; Tue, 18 Nov 2008 12:20:44 +1100 (EST) Received: from postoffice.aconex.com (postoffice.yarra.acx [192.168.102.1]) by postoffice2.aconex.com with ESMTP id EHVLoMDVQJICDpug for ; Tue, 18 Nov 2008 12:20:44 +1100 (EST) Received: from [192.168.5.24] (melho0.aconex.com [203.89.192.141]) by postoffice.aconex.com (Postfix) with ESMTP id 5E44292C312 for ; Tue, 18 Nov 2008 12:10:08 +1100 (EST) X-ASG-Orig-Subj: pcpweb updates Subject: pcpweb updates From: Nathan Scott To: pcp@oss.sgi.com Content-Type: text/plain Date: Tue, 18 Nov 2008 12:18:32 +1100 Message-Id: <1226971112.4409.2.camel@verge.scott.net.au> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: mail.aconex.com[203.89.202.182] X-Barracuda-Start-Time: 1226972330 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -1.52 X-Barracuda-Spam-Status: No, SCORE=-1.52 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests=BSF_RULE7568M X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.10513 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- 0.50 BSF_RULE7568M Custom Rule 7568M X-Virus-Scanned: ClamAV version 0.91.2, clamav-milter version 0.91.2 on oss.sgi.com X-Virus-Status: Clean Changes committed to git://oss.sgi.com/nathans/pcpweb.git index.html | 4 ++-- news.html | 9 ++++++++- source.html | 6 +++--- 3 files changed, 13 insertions(+), 6 deletions(-) commit 8f2eeac91eba39bde5a590215978835e5c5afd1e Merge: f16417c... cf7ad0f... Author: Nathan Scott Date: Tue Nov 18 07:01:33 2008 +1100 Merge branch 'master' of git://oss.sgi.com/nathans/pcpweb commit f16417c2196bc16f4a1796d9a54b79b888197c6f Author: Nathan Scott Date: Mon Nov 17 21:24:31 2008 +1100 Correct development snapshot version number. commit b6c488d3c7ca539e4198e045770c7434f2bc87a7 Author: Nathan Scott Date: Mon Nov 17 21:18:15 2008 +1100 Update website docs - new PMDAs, new PCP development version. commit cf7ad0fd5326bd098c9e623a113663fbd5d3d78a Author: Nathan Scott Date: Fri Nov 7 15:24:20 2008 +1100 Update git tree location - port 8090 being phased out on oss.sgi.com. From nscott@aconex.com Mon Nov 17 17:38:58 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=-1.8 required=5.0 tests=AWL,BAYES_00,J_CHICKENPOX_73 autolearn=no version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAI1cw69010688 for ; Mon, 17 Nov 2008 17:38:58 -0800 X-ASG-Debug-ID: 1226972326-3d7503b20000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from postoffice2.aconex.com (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 3C0311591FFD for ; Mon, 17 Nov 2008 17:38:47 -0800 (PST) Received: from postoffice2.aconex.com (mail.aconex.com [203.89.202.182]) by cuda.sgi.com with ESMTP id 45bHBSbG0VHs46nH for ; Mon, 17 Nov 2008 17:38:47 -0800 (PST) Received: from postoffice.aconex.com (localhost [127.0.0.1]) by postoffice2.aconex.com (Spam Firewall) with ESMTP id 4ACE63A72E7 for ; Tue, 18 Nov 2008 12:20:16 +1100 (EST) Received: from postoffice.aconex.com (postoffice.yarra.acx [192.168.102.1]) by postoffice2.aconex.com with ESMTP id 2NTt2KaBUCYzUz0w for ; Tue, 18 Nov 2008 12:20:16 +1100 (EST) Received: from [192.168.5.24] (melho0.aconex.com [203.89.192.141]) by postoffice.aconex.com (Postfix) with ESMTP id 93BAD92C312 for ; Tue, 18 Nov 2008 12:09:40 +1100 (EST) X-ASG-Orig-Subj: pcp updates (and merge) Subject: pcp updates (and merge) From: Nathan Scott To: pcp@oss.sgi.com Content-Type: text/plain Date: Tue, 18 Nov 2008 12:18:04 +1100 Message-Id: <1226971084.4409.0.camel@verge.scott.net.au> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: mail.aconex.com[203.89.202.182] X-Barracuda-Start-Time: 1226972331 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -0.02 X-Barracuda-Spam-Status: No, SCORE=-0.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests=BSF_RULE7568M, WEIRD_PORT X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.10513 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- 1.50 WEIRD_PORT URI: Uses non-standard port number for HTTP 0.50 BSF_RULE7568M Custom Rule 7568M X-Virus-Scanned: ClamAV version 0.91.2, clamav-milter version 0.91.2 on oss.sgi.com X-Virus-Status: Clean Changes committed to git://oss.sgi.com/nathans/pcp.git VERSION.pcp | 6 build/rpm/pcp.spec.in | 18 debian/changelog | 9 src/libpcp/src/GNUmakefile | 2 src/libpcp/src/connect.c | 12 src/pmdas/GNUmakefile | 8 src/pmdas/ib/GNUmakefile | 79 +++ src/pmdas/ib/Install | 77 +++ src/pmdas/ib/Remove | 40 + src/pmdas/ib/daemon.c | 85 ++++ src/pmdas/ib/dso.c | 43 ++ src/pmdas/ib/help | 169 ++++++++ src/pmdas/ib/ib.c | 894 +++++++++++++++++++++++++++++++++++++++++++ src/pmdas/ib/ibpmda.h | 104 +++++ src/pmdas/ib/pmda.c | 339 ++++++++++++++++ src/pmdas/ib/pmns | 78 +++ src/pmdas/ib/root | 9 src/pmdas/kvm/GNUmakefile | 43 ++ src/pmdas/kvm/Install | 40 + src/pmdas/kvm/Remove | 29 + src/pmdas/kvm/pmdakvm.pl | 112 +++++ src/pmdas/linux/GNUmakefile | 8 src/pmdas/linux/clusters.h | 4 src/pmdas/linux/help | 48 +- src/pmdas/linux/indom.h | 6 src/pmdas/linux/infiniband.c | 573 --------------------------- src/pmdas/linux/infiniband.h | 67 --- src/pmdas/linux/pmda.c | 156 ++----- src/pmns/stdpmid.add-on | 3 src/pmns/stdpmid.pcp | 1 30 files changed, 2281 insertions(+), 781 deletions(-) commit 090ba3670f4aeec3e95a81363c368076beae1d98 Author: Nathan Scott Date: Mon Nov 17 21:28:49 2008 +1100 Bump development snapshot version number. commit a1316d3954d4873daf2b9e1fad504d97b45a4426 Author: Nathan Scott Date: Mon Nov 17 20:06:39 2008 +1100 Pick a unique, unused domain number (93) for the KVM PMDA, and add kvm subdir into the build, only installed on Linux. commit 0d1d1c66c391745143e99894e0335d16c48123c4 Author: Nathan Scott Date: Mon Nov 17 08:57:33 2008 +1100 Initial version of the KVM PMDA, exporting metric values exposed by the Linux kernel virtualisation subsystem, KVM. commit 68c47bb78209783fb875dd0278ede7ce1b84f91e Merge: 7b2760b... 71625d6... Author: Nathan Scott Date: Fri Nov 14 16:08:22 2008 +1100 Merge branch 'dev' of git://oss.sgi.com/pcp/pcp into nathans commit 71625d6aab5ad7d45417d387c351b7a53470f76c Merge: afb8909... ac0f2db... Author: Jonathan Knispel Date: Wed Nov 12 17:34:35 2008 +1100 Merge branch 'nathans' of git://oss.sgi.com:8090/nathans/pcp into dev Signed-off-by: Jonathan Knispel Nathan Scott's branch of PCP on OSS has been updated since Mark cloned it to reconcile changes between the SGI and OSS trees. Merge Nathan's updates back into the SGI reconciliation branch in preparation for pushing it to OSS. commit afb8909814c87aa72658f28bd34f9a4099bc3e41 Author: Jonathan Knispel Date: Fri Oct 31 17:34:52 2008 +1100 Add SGI's InfiniBand PMDA as pmda_ib (domain 91), bump revision to 2.7.8, build 1 Signed-off-by: Jonathan Knispel commit 12070a00274a48dfdf1c4715135584ed898e5472 Author: Jonathan Knispel Date: Fri Oct 31 14:49:40 2008 +1100 Deprecate InfiniBand from pmda_linux Signed-off-by: Jonathan Knispel From nscott@aconex.com Thu Nov 20 23:01:04 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=unavailable version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAL5147K027917 for ; Thu, 20 Nov 2008 23:01:04 -0600 X-ASG-Debug-ID: 1227243662-1275013f0000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from postoffice2.aconex.com (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id CCD2D15BABED for ; Thu, 20 Nov 2008 21:01:02 -0800 (PST) Received: from postoffice2.aconex.com (mail.aconex.com [203.89.202.182]) by cuda.sgi.com with ESMTP id 4RK4md427BBEKeFF for ; Thu, 20 Nov 2008 21:01:02 -0800 (PST) Received: from postoffice.aconex.com (localhost [127.0.0.1]) by postoffice2.aconex.com (Spam Firewall) with ESMTP id 9D6AA3DF1FC for ; Fri, 21 Nov 2008 16:00:59 +1100 (EST) Received: from postoffice.aconex.com (postoffice.yarra.acx [192.168.102.1]) by postoffice2.aconex.com with ESMTP id g2BDJUC0onzCCE1n for ; Fri, 21 Nov 2008 16:00:59 +1100 (EST) Received: from [192.168.5.24] (melho0.aconex.com [203.89.192.141]) by postoffice.aconex.com (Postfix) with ESMTP id 56E7892C2E2 for ; Fri, 21 Nov 2008 15:49:36 +1100 (EST) X-ASG-Orig-Subj: pcp updates Subject: pcp updates From: Nathan Scott To: pcp@oss.sgi.com Content-Type: text/plain Date: Fri, 21 Nov 2008 15:58:45 +1100 Message-Id: <1227243525.4370.52.camel@verge.scott.net.au> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: mail.aconex.com[203.89.202.182] X-Barracuda-Start-Time: 1227243663 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -2.02 X-Barracuda-Spam-Status: No, SCORE=-2.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests= X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.10744 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- Changes committed to git://oss.sgi.com/nathans/pcp.git GNUmakefile | 3 ++- configure.in | 2 +- src/libpcp/src/GNUmakefile | 8 +++++++- 3 files changed, 10 insertions(+), 3 deletions(-) commit 32f96056e7f4e67e94da21f4b5a61f7d05bb749d Author: Nathan Scott Date: Fri Nov 21 08:11:24 2008 +1100 Do not unconditionally add the Infiniband PMDA into the set of DSOs which can be used via local context - only when configured into the build. Noticed while debugging under Win32. commit beb84d862142fc5e250c077a0b7c9bec481a1c41 Author: Nathan Scott Date: Fri Nov 21 08:08:48 2008 +1100 Cleanup PCP_VAR_DIR setting on Win32 so we dont see dup //s on path lookups. From nscott@aconex.com Fri Nov 21 05:48:48 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=unavailable version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mALBmmSA030016 for ; Fri, 21 Nov 2008 05:48:48 -0600 X-ASG-Debug-ID: 1227268127-247e03d20000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from postoffice2.aconex.com (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id F14EA1B88AB0 for ; Fri, 21 Nov 2008 03:48:47 -0800 (PST) Received: from postoffice2.aconex.com (mail.aconex.com [203.89.202.182]) by cuda.sgi.com with ESMTP id uQYTzZcq82KiP7Yp for ; Fri, 21 Nov 2008 03:48:47 -0800 (PST) Received: from postoffice.aconex.com (localhost [127.0.0.1]) by postoffice2.aconex.com (Spam Firewall) with ESMTP id 93F6F3E61F8 for ; Fri, 21 Nov 2008 21:35:32 +1100 (EST) Received: from postoffice.aconex.com (postoffice.yarra.acx [192.168.102.1]) by postoffice2.aconex.com with ESMTP id UHXfNW2KaivZAKRa for ; Fri, 21 Nov 2008 21:35:32 +1100 (EST) Received: from [192.168.0.100] (c220-239-214-222.fernt2.vic.optusnet.com.au [220.239.214.222]) by postoffice.aconex.com (Postfix) with ESMTP id C7A8492C2E2 for ; Fri, 21 Nov 2008 21:24:05 +1100 (EST) X-ASG-Orig-Subj: pcpweb updates Subject: pcpweb updates From: Nathan Scott To: pcp@oss.sgi.com Content-Type: text/plain Date: Fri, 21 Nov 2008 21:33:18 +1100 Message-Id: <1227263598.4269.4.camel@verge.scott.net.au> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: mail.aconex.com[203.89.202.182] X-Barracuda-Start-Time: 1227268127 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -2.02 X-Barracuda-Spam-Status: No, SCORE=-2.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests= X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.10769 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- Changes committed to git://oss.sgi.com/nathans/pcpweb.git faq.html | 2 images/architecture.png |binary images/architecture.svg | 1973 ++++++++++++++++++++++++++++++++++++++++++++++++ win32/pcpcmd.bat | 1 win32/pcpsh.bat | 1 5 files changed, 1976 insertions(+), 1 deletion(-) commit 4d5c003a197acad0739e591bd84b83edfeb18324 Author: Nathan Scott Date: Fri Nov 21 21:32:03 2008 +1100 Change directory before running Win32 shells to start somewhere known good. commit f621c2e3d48b30fbe1fa344b57a13409e6ed9c7c Author: Nathan Scott Date: Fri Nov 21 21:31:33 2008 +1100 A funkier PCP architecture diagram. From nscott@aconex.com Fri Nov 21 05:48:48 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=unavailable version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mALBmmo4030023 for ; Fri, 21 Nov 2008 05:48:48 -0600 X-ASG-Debug-ID: 1227268127-5ea1018d0000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from postoffice2.aconex.com (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 3B3C31B88AB2 for ; Fri, 21 Nov 2008 03:48:47 -0800 (PST) Received: from postoffice2.aconex.com (mail.aconex.com [203.89.202.182]) by cuda.sgi.com with ESMTP id 9AjYn3n9x4fjSwm9 for ; Fri, 21 Nov 2008 03:48:47 -0800 (PST) Received: from postoffice.aconex.com (localhost [127.0.0.1]) by postoffice2.aconex.com (Spam Firewall) with ESMTP id DE4C73E50E4 for ; Fri, 21 Nov 2008 20:17:32 +1100 (EST) Received: from postoffice.aconex.com (postoffice.yarra.acx [192.168.102.1]) by postoffice2.aconex.com with ESMTP id rApuZaPoQaCgicRH for ; Fri, 21 Nov 2008 20:17:32 +1100 (EST) Received: from [192.168.0.100] (c220-239-214-222.fernt2.vic.optusnet.com.au [220.239.214.222]) by postoffice.aconex.com (Postfix) with ESMTP id EE44792C312 for ; Fri, 21 Nov 2008 20:06:06 +1100 (EST) X-ASG-Orig-Subj: pcpqa updates Subject: pcpqa updates From: Nathan Scott To: pcp@oss.sgi.com Content-Type: text/plain Date: Fri, 21 Nov 2008 20:15:18 +1100 Message-Id: <1227258918.4269.0.camel@verge.scott.net.au> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: mail.aconex.com[203.89.202.182] X-Barracuda-Start-Time: 1227268128 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -1.52 X-Barracuda-Spam-Status: No, SCORE=-1.52 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests=BSF_RULE7568M X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.10769 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- 0.50 BSF_RULE7568M Custom Rule 7568M Changes committed to git://oss.sgi.com/nathans/pcpqa.git 348 | 117 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 348.out | 33 ++++++++++++++++++ group | 3 + owner | 1 4 files changed, 154 insertions(+) commit b0a572675159103825f962c3087097077c231a2e Author: Nathan Scott Date: Fri Nov 21 08:16:42 2008 +1100 Add test 348 into owner/group files. commit 6cacf9bd4d37e9cca7a258609a8a2c429bcb0cd9 Author: Nathan Scott Date: Fri Nov 21 08:16:06 2008 +1100 Add a QA test to exercise the KVM PMDA - only run on Linux and only if the KVM stats are being exported though sysfs. Exercises the Install/Remove logic, and verifies that the values match up. From kenj@internode.on.net Mon Nov 24 12:09:15 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=unavailable version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAOI9DTs028812 for ; Mon, 24 Nov 2008 12:09:15 -0600 X-ASG-Debug-ID: 1227550151-1e8502850000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from mail.internode.on.net (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 289E01BDF3C7 for ; Mon, 24 Nov 2008 10:09:11 -0800 (PST) Received: from mail.internode.on.net (bld-mail05.adl2.internode.on.net [203.16.214.69]) by cuda.sgi.com with ESMTP id JzGlHGjCdOUO1ODH for ; Mon, 24 Nov 2008 10:09:11 -0800 (PST) Received: from [192.168.1.100] (unverified [118.208.222.17]) by mail.internode.on.net (SurgeMail 3.8f2) with ESMTP id 34943227-1927428 for multiple; Tue, 25 Nov 2008 04:23:20 +1030 (CDT) X-ASG-Orig-Subj: Re: Daylight savings bug Subject: Re: Daylight savings bug From: Ken McDonell Reply-To: kenj@internode.on.net To: Nathan Scott Cc: pcp@oss.sgi.com In-Reply-To: <1223438632.4494.17.camel@verge.scott.net.au> References: <1223438632.4494.17.camel@verge.scott.net.au> Content-Type: text/plain Date: Tue, 25 Nov 2008 04:53:05 +1100 Message-Id: <1227549185.10096.13.camel@bozo> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: bld-mail05.adl2.internode.on.net[203.16.214.69] X-Barracuda-Start-Time: 1227550153 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -2.02 X-Barracuda-Spam-Status: No, SCORE=-2.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests= X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11051 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- I may have a handle on this. Nathan, can you please check the following for me. 1. the summary archive for 20081004 is ok. 2 the summary archive for 20081005 is missing. 3. now can you check the summary archive for 20081006 ... I'm interested in the actual dates for the label and the log records in that archive ... I suspect it is really data for 20081005. 4. and if that is the case, I suspect you also have the component archives for 20081006 that have not been merged (although these may well have been culled by now). On Wed, 2008-10-08 at 15:03 +1100, Nathan Scott wrote: > Hi all, > > Just wanted to send a note - I've not had an opportunity > to look into this, and wont soon - perhaps others can in > the meantime. > > On Sunday, we had a daylight savings change (Melbourne, > Victoria, Australia), clocks going forward one hour at 2am. > Somehow the daily PCP log rotation scripts have managed > to delete all of the PCP archives for that day, I presume > as part of the daily log rotation. We use the recommended > daily log rotation regime, with daily rotation at 00:10 as > per the man pages, driven by cron. > > This happened on every single host I looked at so far, so > not some wierd race condition. Presumably something in the > logic that does the log merge of all individual logs for 1 > day, and then unlinks the old ones has resulted in no "new" > merged archive (how!?) and yet still removed the "old" ones. > Looking back, I don't think this is the first time it has > happened. Argh! > > cheers. > > -- > Nathan > > From nscott@aconex.com Mon Nov 24 16:58:49 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=unavailable version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAOMwmAm017593 for ; Mon, 24 Nov 2008 16:58:49 -0600 X-ASG-Debug-ID: 1227567526-2f9f005d0000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from postoffice2.aconex.com (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 110171B89C86 for ; Mon, 24 Nov 2008 14:58:46 -0800 (PST) Received: from postoffice2.aconex.com (mail.aconex.com [203.89.202.182]) by cuda.sgi.com with ESMTP id w2dTas5arpcKkTZx for ; Mon, 24 Nov 2008 14:58:46 -0800 (PST) Received: from postoffice.aconex.com (localhost [127.0.0.1]) by postoffice2.aconex.com (Spam Firewall) with ESMTP id 43D9066B62E; Tue, 25 Nov 2008 09:40:38 +1100 (EST) Received: from postoffice.aconex.com (postoffice.yarra.acx [192.168.102.1]) by postoffice2.aconex.com with ESMTP id KTrMCjAsmSzLY4fb; Tue, 25 Nov 2008 09:40:38 +1100 (EST) Received: from [192.168.5.24] (melho0.aconex.com [203.89.192.141]) by postoffice.aconex.com (Postfix) with ESMTP id 33D6A92C2E2; Tue, 25 Nov 2008 09:40:38 +1100 (EST) X-ASG-Orig-Subj: Re: Daylight savings bug Subject: Re: Daylight savings bug From: Nathan Scott To: kenj@internode.on.net Cc: pcp@oss.sgi.com In-Reply-To: <1227549185.10096.13.camel@bozo> References: <1223438632.4494.17.camel@verge.scott.net.au> <1227549185.10096.13.camel@bozo> Content-Type: text/plain Date: Tue, 25 Nov 2008 09:38:19 +1100 Message-Id: <1227566299.4557.10.camel@verge.scott.net.au> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: mail.aconex.com[203.89.202.182] X-Barracuda-Start-Time: 1227567528 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -2.02 X-Barracuda-Spam-Status: No, SCORE=-2.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests= X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11068 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- On Tue, 2008-11-25 at 04:53 +1100, Ken McDonell wrote: > I may have a handle on this. > > Nathan, can you please check the following for me. > > 1. the summary archive for 20081004 is ok. Yep. > 2 the summary archive for 20081005 is missing. Yep. > 3. now can you check the summary archive for 20081006 ... I'm interested > in the actual dates for the label and the log records in that > archive ... I suspect it is really data for 20081005. See below, appears to really be data for 06 AFAICT. > 4. and if that is the case, I suspect you also have the component > archives for 20081006 that have not been merged (although these may well > have been culled by now). Yeah, no component archives left anywhere now that I can see. $ pmdumplog -l 20081004 Log Label (Log Format Version 2) Performance metrics from host app4 commencing Sat Oct 4 00:13:41.741 2008 ending Sun Oct 5 00:13:26.753 2008 $ pmdumplog -l 20081006 Log Label (Log Format Version 2) Performance metrics from host app4 commencing Mon Oct 6 00:12:14.239 2008 ending Tue Oct 7 00:13:35.340 2008 All options I tried to pmdumplog give timestamps relative to start of the day (and all appeared to be from the start of 06). Trying to get more info with -v was unhelpful - a bug lurking here too perhaps? All archives produce this: $ pmdumplog -v 20081006.0 Raw dump of physical archive file "20081006.0" ... Dump ... record len: -2080374784 @ offset: 0 Trailer botch: 36046160 != -2080374784 Is there a better way to dump out the timestamps than with pmdumplog? (or an option I've missed? -Dsomething?) thanks! -- Nathan From nscott@aconex.com Mon Nov 24 18:09:50 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=unavailable version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAP09nLk024361 for ; Mon, 24 Nov 2008 18:09:50 -0600 X-ASG-Debug-ID: 1227571787-2f7101410000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from postoffice2.aconex.com (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id D2A4515E3179 for ; Mon, 24 Nov 2008 16:09:47 -0800 (PST) Received: from postoffice2.aconex.com (mail.aconex.com [203.89.202.182]) by cuda.sgi.com with ESMTP id NXlBYfXuB3ZpIf6H for ; Mon, 24 Nov 2008 16:09:47 -0800 (PST) Received: from postoffice.aconex.com (localhost [127.0.0.1]) by postoffice2.aconex.com (Spam Firewall) with ESMTP id 761A5400ED2; Tue, 25 Nov 2008 10:50:58 +1100 (EST) Received: from postoffice.aconex.com (postoffice.yarra.acx [192.168.102.1]) by postoffice2.aconex.com with ESMTP id Cztw7COFF9OynKCA; Tue, 25 Nov 2008 10:50:58 +1100 (EST) Received: from [192.168.5.24] (melho0.aconex.com [203.89.192.141]) by postoffice.aconex.com (Postfix) with ESMTP id 2152E92C313; Tue, 25 Nov 2008 10:50:58 +1100 (EST) X-ASG-Orig-Subj: Re: Daylight savings bug Subject: Re: Daylight savings bug From: Nathan Scott To: kenj@internode.on.net Cc: pcp@oss.sgi.com In-Reply-To: <1227569919.10096.29.camel@bozo> References: <1223438632.4494.17.camel@verge.scott.net.au> <1227549185.10096.13.camel@bozo> <1227566299.4557.10.camel@verge.scott.net.au> <1227569919.10096.29.camel@bozo> Content-Type: multipart/mixed; boundary="=-LmqgpZV5ZKj34OGT3ltO" Date: Tue, 25 Nov 2008 10:48:39 +1100 Message-Id: <1227570519.4557.24.camel@verge.scott.net.au> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 X-Barracuda-Connect: mail.aconex.com[203.89.202.182] X-Barracuda-Start-Time: 1227571788 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -2.02 X-Barracuda-Spam-Status: No, SCORE=-2.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests= X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11071 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- --=-LmqgpZV5ZKj34OGT3ltO Content-Type: text/plain Content-Transfer-Encoding: 7bit On Tue, 2008-11-25 at 10:38 +1100, Ken McDonell wrote: > Hmmm. > > Does not disprove my hypothesis, but does not prove it either. > > 20081006 is a bit odd. If pmdumplog -l reports an end time then the > archive is not really busted. But pmdumplog -v is a strange beast, that > looks like a pmdumplog bug (sigh). > > Just try pmdumplog 20081006 | head -100 ... the timestamps appear in the > pmResult dumps. I've attached two dump heads - one pmdumplog with no options, the other with -Dpdu (this one shows the raw timestamps). > But 20081006 looks like it is for the correct day which undermines my > theory. > > I believe the problem is in pmdate -1d when you're within an hour of > midnight and there was a DST adjustment in the previous 23-25 hrs ... > there is no good answer here, pmdate cannot produce the 100% right > answer no matter what. Yep, sounds entirely feasible. > I'm looking at reworking pmlogger_daily to remove all dependence on > pmdate, and this will give me a chance to re-instate the -o option ... > actually new behaviour to merge _all_ archives, not just yesterday's and > -o for the old (current) behaviour, so all of pmlogger_daily QA will > work again ... OK, great! cheers. -- Nathan --=-LmqgpZV5ZKj34OGT3ltO Content-Disposition: attachment; filename=dumplog.head Content-Type: text/plain; name=dumplog.head; charset=utf-8 Content-Transfer-Encoding: 7bit 00:12:14.239 2.3.3 (pmcd.pmlogger.host): inst [23112 or "23112"] value "web2.mel.acx" 2.3.0 (pmcd.pmlogger.port): inst [23112 or "23112"] value 4336 2.3.2 (pmcd.pmlogger.archive): inst [23112 or "23112"] value "/var/log/pcp/pmlogger/app4/20081006.00.10" 00:12:14.242 60.18.7 (hinv.machine): value "linux" 60.18.6 (hinv.map.cpu_num): inst [0 or "cpu0"] value 0 inst [1 or "cpu1"] value 1 inst [2 or "cpu2"] value 2 inst [3 or "cpu3"] value 3 inst [4 or "cpu4"] value 4 inst [5 or "cpu5"] value 5 inst [6 or "cpu6"] value 6 inst [7 or "cpu7"] value 7 60.18.5 (hinv.cpu.bogomips): inst [0 or "cpu0"] value 5656.3901 inst [1 or "cpu1"] value 5652.7598 inst [2 or "cpu2"] value 5652.5298 inst [3 or "cpu3"] value 5652.52 inst [4 or "cpu4"] value 5652.52 inst [5 or "cpu5"] value 5652.54 inst [6 or "cpu6"] value 5652.6699 inst [7 or "cpu7"] value 5652.5298 60.18.4 (hinv.cpu.cache): inst [0 or "cpu0"] value 6144 inst [1 or "cpu1"] value 6144 inst [2 or "cpu2"] value 6144 inst [3 or "cpu3"] value 6144 inst [4 or "cpu4"] value 6144 inst [5 or "cpu5"] value 6144 inst [6 or "cpu6"] value 6144 inst [7 or "cpu7"] value 6144 60.18.3 (hinv.cpu.stepping): inst [0 or "cpu0"] value "6" inst [1 or "cpu1"] value "6" inst [2 or "cpu2"] value "6" inst [3 or "cpu3"] value "6" inst [4 or "cpu4"] value "6" inst [5 or "cpu5"] value "6" inst [6 or "cpu6"] value "6" inst [7 or "cpu7"] value "6" 60.18.2 (hinv.cpu.model): inst [0 or "cpu0"] value "23" inst [1 or "cpu1"] value "23" inst [2 or "cpu2"] value "23" inst [3 or "cpu3"] value "23" inst [4 or "cpu4"] value "23" inst [5 or "cpu5"] value "23" inst [6 or "cpu6"] value "23" inst [7 or "cpu7"] value "23" 60.18.1 (hinv.cpu.vendor): inst [0 or "cpu0"] value "GenuineIntel" inst [1 or "cpu1"] value "GenuineIntel" inst [2 or "cpu2"] value "GenuineIntel" inst [3 or "cpu3"] value "GenuineIntel" inst [4 or "cpu4"] value "GenuineIntel" inst [5 or "cpu5"] value "GenuineIntel" inst [6 or "cpu6"] value "GenuineIntel" inst [7 or "cpu7"] value "GenuineIntel" 60.18.0 (hinv.cpu.clock): inst [0 or "cpu0"] value 2826.2849 inst [1 or "cpu1"] value 2826.2849 inst [2 or "cpu2"] value 2826.2849 inst [3 or "cpu3"] value 2826.2849 inst [4 or "cpu4"] value 2826.2849 inst [5 or "cpu5"] value 2826.2849 inst [6 or "cpu6"] value 2826.2849 inst [7 or "cpu7"] value 2826.2849 60.15.0 (hinv.map.scsi): inst [0 or "scsi0:0:32:0 Enclosure"] value "Unknown SCSI device" inst [1 or "scsi0:2:0:0 Direct-Access"] value "sda" inst [2 or "scsi0:2:1:0 Direct-Access"] value "sdb" inst [3 or "scsi1:0:0:0 CD-ROM"] value "scd0" inst [4 or "scsi2:0:0:0 Direct-Access"] value "sdc" 60.12.7 (kernel.uname.distro): value "Red Hat Enterprise Linux Server release 5.2 (Tikanga)" 60.12.4 (kernel.uname.nodename): value "app4.mel.acx" 60.12.3 (kernel.uname.machine): value "x86_64" 60.12.2 (kernel.uname.sysname): value "Linux" 60.12.1 (kernel.uname.version): value "#1 SMP Fri Jun 20 02:36:06 EDT 2008" 60.12.0 (kernel.uname.release): value "2.6.18-92.1.6.el5" 60.5.1 (filesys.capacity): inst [0 or "/dev/root"] value 6092360 inst [1 or "/dev"] value 8216116 inst [2 or "/dev/sda13"] value 4061540 inst [3 or "/dev/sda3"] value 29458408 inst [4 or "/dev/sda8"] value 6092360 inst [5 or "/dev/sda1"] value 101086 inst [6 or "/dev/sda6"] value 10317828 inst [7 or "/dev/sda14"] value 163875760 inst [8 or "/dev/sda11"] value 4127076 inst [9 or "/dev/sdb1"] value 575661796 60.5.0 (hinv.nfilesys): value 10 60.1.11 (hinv.pagesize): value 4096 60.1.9 (hinv.physmem): value 16047 60.0.33 (hinv.ndisk): value 4 60.0.32 (hinv.ncpu): value 8 00:12:14.274 60.20.2 (mem.slabinfo.objects.size): inst [33 or "ext3_inode_cache"] value 760 --=-LmqgpZV5ZKj34OGT3ltO Content-Disposition: attachment; filename=dumplog.head2 Content-Type: text/plain; name=dumplog.head2; charset=utf-8 Content-Transfer-Encoding: 7bit pmResult dump from 0x60fb50 timestamp: 1223298815.340077 00:13:35.340 numpmid: 2 60.7.61 (nfs3.client.reqs): numval: 22 valfmt: 0 vlist[]: inst [0 or ???] value 0 inst [1 or ???] value 4207671 inst [2 or ???] value 598580 inst [3 or ???] value 4410068 inst [4 or ???] value 4061564 inst [5 or ???] value 0 inst [6 or ???] value 23262760 inst [7 or ???] value 10763577 inst [8 or ???] value 468260 inst [9 or ???] value 44225 inst [10 or ???] value 0 inst [11 or ???] value 0 inst [12 or ???] value 263079 inst [13 or ???] value 75566 inst [14 or ???] value 234420 inst [15 or ???] value 0 inst [16 or ???] value 427 inst [17 or ???] value 31331 inst [18 or ???] value 172 inst [19 or ???] value 24 inst [20 or ???] value 0 inst [21 or ???] value 345998 60.7.60 (nfs3.client.calls): numval: 1 valfmt: 0 vlist[]: value 48767722 __pmLogRead timestamp=00:13:35.340 0x61000c ... 0x6100f0 000: ff0eea48 6d300500 2000000 3d1c000f 16000000 0 0 0 008: 1000000 37344000 2000000 34220900 3000000 d44a4300 4000000 7cf93d00 016: 5000000 0 6000000 28f66201 7000000 393da400 8000000 24250700 024: 9000000 c1ac0000 a000000 0 b000000 0 c000000 a7030400 032: d000000 2e270100 e000000 b4930300 f000000 0 10000000 ab010000 040: 11000000 637a0000 12000000 ac000000 13000000 18000000 14000000 0 048: 15000000 8e470500 3c1c000f 1000000 0 ffffffff ea22e802 pmResult dump from 0x60fc30 timestamp: 1223212334.239894 00:12:14.239 numpmid: 3 2.3.3 (pmcd.pmlogger.host): numval: 1 valfmt: 1 vlist[]: inst [23112 or "23112"] value "web2.mel.acx" 2.3.0 (pmcd.pmlogger.port): numval: 1 valfmt: 0 vlist[]: inst [23112 or "23112"] value 4336 2.3.2 (pmcd.pmlogger.archive): numval: 1 valfmt: 1 vlist[]: inst [23112 or "23112"] value "/var/log/pcp/pmlogger/app4/20081006.00.10" __pmLogRead timestamp=00:12:14.239 0x61100c ... 0x6110a0 000: 2ebde848 16a90300 3000000 30c8000 1000000 1000000 485a0000 15000000 008: c8000 1000000 0 485a0000 f0100000 20c8000 1000000 1000000 016: 485a0000 1a000000 6000011 32626577 6c656d2e 7863612e 7e7e7e00 600002e 024: 7261762f 676f6c2f 7063702f 6c6d702f 6567676f 70612f72 322f3470 31383030 032: 2e363030 312e3030 7e7e0030 pmResult dump from 0x60fc30 timestamp: 1223212334.242047 00:12:14.242 numpmid: 21 60.18.7 (hinv.machine): numval: 1 valfmt: 1 vlist[]: value "linux" 60.18.6 (hinv.map.cpu_num): numval: 8 valfmt: 0 vlist[]: inst [0 or ???] value 0 inst [1 or ???] value 1 inst [2 or ???] value 2 inst [3 or ???] value 3 inst [4 or ???] value 4 inst [5 or ???] value 5 inst [6 or ???] value 6 inst [7 or ???] value 7 60.18.5 (hinv.cpu.bogomips): numval: 8 valfmt: 1 vlist[]: inst [0 or ???] value 5656.3901 inst [1 or ???] value 5652.7598 inst [2 or ???] value 5652.5298 inst [3 or ???] value 5652.52 inst [4 or ???] value 5652.52 inst [5 or ???] value 5652.54 inst [6 or ???] value 5652.6699 inst [7 or ???] value 5652.5298 60.18.4 (hinv.cpu.cache): numval: 8 valfmt: 0 vlist[]: inst [0 or ???] value 6144 inst [1 or ???] value 6144 inst [2 or ???] value 6144 inst [3 or ???] value 6144 inst [4 or ???] value 6144 inst [5 or ???] value 6144 inst [6 or ???] value 6144 inst [7 or ???] value 6144 60.18.3 (hinv.cpu.stepping): numval: 8 valfmt: 1 vlist[]: inst [0 or ???] value "6" inst [1 or ???] value "6" inst [2 or ???] value "6" inst [3 or ???] value "6" inst [4 or ???] value "6" inst [5 or ???] value "6" inst [6 or ???] value "6" inst [7 or ???] value "6" 60.18.2 (hinv.cpu.model): numval: 8 valfmt: 1 vlist[]: inst [0 or ???] value "23" inst [1 or ???] value "23" inst [2 or ???] value "23" inst [3 or ???] value "23" inst [4 or ???] value "23" inst [5 or ???] value "23" inst [6 or ???] value "23" inst [7 or ???] value "23" 60.18.1 (hinv.cpu.vendor): numval: 8 valfmt: 1 vlist[]: inst [0 or ???] value "GenuineIntel" inst [1 or ???] value "GenuineIntel" inst [2 or ???] value "GenuineIntel" inst [3 or ???] value "GenuineIntel" --=-LmqgpZV5ZKj34OGT3ltO-- From nscott@aconex.com Mon Nov 24 18:18:50 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: * X-Spam-Status: No, score=2.0 required=5.0 tests=URIBL_BLACK autolearn=no version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAP0InuL024782 for ; Mon, 24 Nov 2008 18:18:50 -0600 X-ASG-Debug-ID: 1227572327-2f78016b0000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from postoffice2.aconex.com (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 7CF841B89C9B for ; Mon, 24 Nov 2008 16:18:47 -0800 (PST) Received: from postoffice2.aconex.com (mail.aconex.com [203.89.202.182]) by cuda.sgi.com with ESMTP id 5uj8CzAkuu2j8vnx for ; Mon, 24 Nov 2008 16:18:47 -0800 (PST) Received: from postoffice.aconex.com (localhost [127.0.0.1]) by postoffice2.aconex.com (Spam Firewall) with ESMTP id 1636B401104 for ; Tue, 25 Nov 2008 10:29:05 +1100 (EST) Received: from postoffice.aconex.com (postoffice.yarra.acx [192.168.102.1]) by postoffice2.aconex.com with ESMTP id Qd4AssgMjKW1oqij for ; Tue, 25 Nov 2008 10:29:05 +1100 (EST) Received: from [192.168.5.24] (melho0.aconex.com [203.89.192.141]) by postoffice.aconex.com (Postfix) with ESMTP id EDD5B92C312 for ; Tue, 25 Nov 2008 10:29:04 +1100 (EST) X-ASG-Orig-Subj: pcp updates Subject: pcp updates From: Nathan Scott To: pcp@oss.sgi.com Content-Type: text/plain Date: Tue, 25 Nov 2008 10:26:46 +1100 Message-Id: <1227569206.4557.18.camel@verge.scott.net.au> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: mail.aconex.com[203.89.202.182] X-Barracuda-Start-Time: 1227572328 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -1.52 X-Barracuda-Spam-Status: No, SCORE=-1.52 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests=BSF_RULE7568M X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11071 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- 0.50 BSF_RULE7568M Custom Rule 7568M Changes committed to git://oss.sgi.com/nathans/pcp.git configure.in | 7 + src/include/builddefs.in | 13 +-- src/pmdas/dbping/GNUmakefile | 18 ++++ src/pmdas/dbping/README | 51 ------------- src/pmdas/dbping/dbprobe.pl | 27 ++++-- src/pmdas/dbping/pmdadbping.pl | 93 ++++++++++++++++++++--- src/pmdas/kvm/GNUmakefile | 14 +++ src/pmdas/kvm/pmdakvm.pl | 12 +-- src/pmdas/memcache/GNUmakefile | 19 +++- src/pmdas/memcache/README | 52 ------------- src/pmdas/memcache/pmdamemcache.pl | 137 +++++++++++++++++++++++++---------- src/pmdas/mysql/GNUmakefile | 17 +++- src/pmdas/mysql/pmdamysql.pl | 81 ++++++++++++++++++++ src/pmdas/systemtap/GNUmakefile | 16 +++- src/pmdas/systemtap/pmdasystemtap.pl | 82 ++++++++++++++++++++ src/pmdas/vmware/GNUmakefile | 14 +++ src/pmdas/vmware/pmdavmware.pl | 77 +++++++++++++++++++ src/pmns/stdpmid.add-on | 2 src/pmns/stdpmid.pcp | 2 19 files changed, 540 insertions(+), 194 deletions(-) commit e0ed285f0b6d23878a349b493482aa5fbf682434 Author: Nathan Scott Date: Tue Nov 25 10:24:47 2008 +1100 Update the makefiles for each of the Perl PMDAs to be able to generate man pages for each, and install them appropriately during the build. Also added man page POD docs to the MySQL PMDA, which was the last one missing a man page. commit 6ee1ccbf34b3e1e83c00374d4801e7c466efc780 Author: Nathan Scott Date: Tue Nov 25 10:21:16 2008 +1100 Add configure magic to find the pod2man utility, and some makefile macros to use it. This gives us the ability to generate installable man pages from inline "pod" docs in the Perl PMDAs. commit 7552b66acae654c9e1fe067622d7870dc6b536e7 Author: Nathan Scott Date: Mon Nov 24 21:51:41 2008 +1100 Update the standard domain numbers to avoid conflicts once more. Shift the Lustre PMDA to slot 93, to match the SGI tree and avoid the memcache PMDA. Also shunt the KVM PMDA down to slot 95, since it'd inadvertantly been added in at 93 as well. What a zoo! :) commit 1a10e27ed8d33c60012a2b0dab22d3c680de0385 Author: Nathan Scott Date: Mon Nov 24 10:03:17 2008 +1100 Manual pages for many of the Perl PMDAs. We should be able to invoke pod2man to generate those for installation, but thats a job for another day (as are the last couple of remaining Perl PMDAs needing better docs). From markgw@sgi.com Mon Nov 24 18:20:27 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=unavailable version=3.3.0-rupdated Received: from relay.sgi.com (relay1.corp.sgi.com [137.38.102.111]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAP0KRM1024920 for ; Mon, 24 Nov 2008 18:20:27 -0600 Received: from larry.melbourne.sgi.com (larry.melbourne.sgi.com [134.14.52.130]) by relay1.corp.sgi.com (Postfix) with SMTP id 8EAC38F8052; Mon, 24 Nov 2008 16:20:22 -0800 (PST) Received: from [134.14.55.22] (dhcp22.melbourne.sgi.com [134.14.55.22]) by larry.melbourne.sgi.com (950413.SGI.8.6.12/950213.SGI.AUTOCF) via ESMTP id LAA26748; Tue, 25 Nov 2008 11:20:19 +1100 Message-ID: <492B44C3.7000000@sgi.com> Date: Tue, 25 Nov 2008 11:20:19 +1100 From: Mark Goodwin Reply-To: markgw@sgi.com Organization: SGI Engineering User-Agent: Thunderbird 2.0.0.18 (Windows/20081105) MIME-Version: 1.0 To: Nathan Scott CC: kenj@internode.on.net, pcp@oss.sgi.com Subject: Re: [pcp] Daylight savings bug References: <1223438632.4494.17.camel@verge.scott.net.au> <1227549185.10096.13.camel@bozo> <1227566299.4557.10.camel@verge.scott.net.au> <1227569919.10096.29.camel@bozo> <1227570519.4557.24.camel@verge.scott.net.au> In-Reply-To: <1227570519.4557.24.camel@verge.scott.net.au> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Nathan Scott wrote: > On Tue, 2008-11-25 at 10:38 +1100, Ken McDonell wrote: >> I believe the problem is in pmdate -1d when you're within an hour of >> midnight and there was a DST adjustment in the previous 23-25 hrs ... >> there is no good answer here, pmdate cannot produce the 100% right >> answer no matter what. > pmdate should return an error when the date/time it is being asked to report does not exist in the local timezone. (sounds like SciFi, but it's a real "temporal discontinuity"). The error would have to be handled all the way out to pmlogger_daily et al. Cheers -- Mark From kenj@internode.on.net Mon Nov 24 18:44:48 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=unavailable version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAP0imCe026224 for ; Mon, 24 Nov 2008 18:44:48 -0600 X-ASG-Debug-ID: 1227573886-2f7401d70000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from mail.internode.on.net (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 0DFBB15E2564 for ; Mon, 24 Nov 2008 16:44:46 -0800 (PST) Received: from mail.internode.on.net (bld-mail08.adl2.internode.on.net [203.16.214.72]) by cuda.sgi.com with ESMTP id xPokSmIARB2vNsmg for ; Mon, 24 Nov 2008 16:44:46 -0800 (PST) Received: from [192.168.1.100] (unverified [118.208.222.17]) by mail.internode.on.net (SurgeMail 3.8f2) with ESMTP id 35272502-1927428 for multiple; Tue, 25 Nov 2008 11:11:33 +1030 (CDT) X-ASG-Orig-Subj: Re: [pcp] Daylight savings bug Subject: Re: [pcp] Daylight savings bug From: Ken McDonell Reply-To: kenj@internode.on.net To: markgw@sgi.com Cc: Nathan Scott , pcp@oss.sgi.com In-Reply-To: <492B44C3.7000000@sgi.com> References: <1223438632.4494.17.camel@verge.scott.net.au> <1227549185.10096.13.camel@bozo> <1227566299.4557.10.camel@verge.scott.net.au> <1227569919.10096.29.camel@bozo> <1227570519.4557.24.camel@verge.scott.net.au> <492B44C3.7000000@sgi.com> Content-Type: text/plain Date: Tue, 25 Nov 2008 11:41:17 +1100 Message-Id: <1227573677.10096.39.camel@bozo> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: bld-mail08.adl2.internode.on.net[203.16.214.72] X-Barracuda-Start-Time: 1227573888 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -2.02 X-Barracuda-Spam-Status: No, SCORE=-2.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests= X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11073 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- The time may well exist, perhaps twice! Consider 00:10 on the morning AFTER a DST change. What is the time (or more specifically, the date) now - 1day? What is the time now - 24hr? utilities like find -mtime -1 don't return an error if there happens to have been a DST change in the last 23-25 hrs. The use of pmdate -anything is suspect in pmlogger_daily ... imho it is cleaner to remove the use of pmdate in that context and leave pmdate to be quietly ignorant and confused in this corner case. On Tue, 2008-11-25 at 11:20 +1100, Mark Goodwin wrote: > > Nathan Scott wrote: > > On Tue, 2008-11-25 at 10:38 +1100, Ken McDonell wrote: > >> I believe the problem is in pmdate -1d when you're within an hour of > >> midnight and there was a DST adjustment in the previous 23-25 hrs ... > >> there is no good answer here, pmdate cannot produce the 100% right > >> answer no matter what. > > > > pmdate should return an error when the date/time it is being asked > to report does not exist in the local timezone. (sounds like SciFi, > but it's a real "temporal discontinuity"). The error would have > to be handled all the way out to pmlogger_daily et al. > > Cheers > -- Mark From kenj@internode.on.net Mon Nov 24 19:24:12 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=unavailable version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAP1OBCF028347 for ; Mon, 24 Nov 2008 19:24:12 -0600 X-ASG-Debug-ID: 1227576249-2f7902830000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from mail.internode.on.net (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 340111BDF687 for ; Mon, 24 Nov 2008 17:24:09 -0800 (PST) Received: from mail.internode.on.net (bld-mail08.adl2.internode.on.net [203.16.214.72]) by cuda.sgi.com with ESMTP id JqKJQVvtEDDYXxaM for ; Mon, 24 Nov 2008 17:24:09 -0800 (PST) Received: from [192.168.1.100] (unverified [118.208.222.17]) by mail.internode.on.net (SurgeMail 3.8f2) with ESMTP id 35276729-1927428 for ; Tue, 25 Nov 2008 11:54:07 +1030 (CDT) X-ASG-Orig-Subj: Re: Daylight savings bug - fix for pmdumplog -v Subject: Re: Daylight savings bug - fix for pmdumplog -v From: Ken McDonell Reply-To: kenj@internode.on.net To: pcp@oss.sgi.com In-Reply-To: <1227573397.4557.34.camel@verge.scott.net.au> References: <1223438632.4494.17.camel@verge.scott.net.au> <1227549185.10096.13.camel@bozo> <1227566299.4557.10.camel@verge.scott.net.au> <1227571952.10096.34.camel@bozo> <1227573397.4557.34.camel@verge.scott.net.au> Content-Type: text/plain Date: Tue, 25 Nov 2008 12:23:52 +1100 Message-Id: <1227576232.10096.50.camel@bozo> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: bld-mail08.adl2.internode.on.net[203.16.214.72] X-Barracuda-Start-Time: 1227576251 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0001 1.0000 -2.0201 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -2.02 X-Barracuda-Spam-Status: No, SCORE=-2.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests= X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11076 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- Any clues how to reactivate accounts on oss.sgi.com? All of the contact/help links on oss.sgi.com seem to lead into marketing land. On Tue, 2008-11-25 at 11:36 +1100, Nathan Scott wrote: > On Tue, 2008-11-25 at 11:12 +1100, Ken McDonell wrote: > > On Tue, 2008-11-25 at 09:38 +1100, Nathan Scott wrote: > > > ... > > I have a patch to fix this (missing endian conversion from the year > > dot). > > > > Patch is attached, which raises the question of the "correct" protocol > > for bozos outside the sgi-aconex consortium to get pcp patches into the > > one true official tree ... guidance? > > 1. Clone yerself a git tree from git://oss.sgi.com/pcp/pcp.git > 2. Create a branch for development > 3. Code, test, commit patches locally > 4. Push changes out to a visible server > 5. Send mail to pcp@oss.sgi.com with updates > > Probably best if you re-activate your account on oss.sgi.com, > unless you have a local server you can use for exporting your > git tree to the world? > > Then, people (esp. Jonathan / myself) can pull in your changes > regularly and I'd expect 'em to appear in the next release. > > cheers. > > -- > Nathan > From nscott@aconex.com Mon Nov 24 19:28:48 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=unavailable version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAP1SmcB028537 for ; Mon, 24 Nov 2008 19:28:48 -0600 X-ASG-Debug-ID: 1227576526-2f7902950000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from postoffice2.aconex.com (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 8CB061BDF6ED for ; Mon, 24 Nov 2008 17:28:47 -0800 (PST) Received: from postoffice2.aconex.com (mail.aconex.com [203.89.202.182]) by cuda.sgi.com with ESMTP id 1r3KFRYRftW9GeCQ for ; Mon, 24 Nov 2008 17:28:47 -0800 (PST) Received: from postoffice.aconex.com (localhost [127.0.0.1]) by postoffice2.aconex.com (Spam Firewall) with ESMTP id 3A14540166C; Tue, 25 Nov 2008 11:38:56 +1100 (EST) Received: from postoffice.aconex.com (postoffice.yarra.acx [192.168.102.1]) by postoffice2.aconex.com with ESMTP id UA94OXUwqfMoo3yQ; Tue, 25 Nov 2008 11:38:56 +1100 (EST) Received: from [192.168.5.24] (melho0.aconex.com [203.89.192.141]) by postoffice.aconex.com (Postfix) with ESMTP id 209E792C2E2; Tue, 25 Nov 2008 11:38:56 +1100 (EST) X-ASG-Orig-Subj: Re: Daylight savings bug - fix for pmdumplog -v Subject: Re: Daylight savings bug - fix for pmdumplog -v From: Nathan Scott To: kenj@internode.on.net Cc: pcp@oss.sgi.com In-Reply-To: <1227571952.10096.34.camel@bozo> References: <1223438632.4494.17.camel@verge.scott.net.au> <1227549185.10096.13.camel@bozo> <1227566299.4557.10.camel@verge.scott.net.au> <1227571952.10096.34.camel@bozo> Content-Type: text/plain Date: Tue, 25 Nov 2008 11:36:37 +1100 Message-Id: <1227573397.4557.34.camel@verge.scott.net.au> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: mail.aconex.com[203.89.202.182] X-Barracuda-Start-Time: 1227576527 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -2.02 X-Barracuda-Spam-Status: No, SCORE=-2.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests= X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11076 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- On Tue, 2008-11-25 at 11:12 +1100, Ken McDonell wrote: > On Tue, 2008-11-25 at 09:38 +1100, Nathan Scott wrote: > > ... > I have a patch to fix this (missing endian conversion from the year > dot). > > Patch is attached, which raises the question of the "correct" protocol > for bozos outside the sgi-aconex consortium to get pcp patches into the > one true official tree ... guidance? 1. Clone yerself a git tree from git://oss.sgi.com/pcp/pcp.git 2. Create a branch for development 3. Code, test, commit patches locally 4. Push changes out to a visible server 5. Send mail to pcp@oss.sgi.com with updates Probably best if you re-activate your account on oss.sgi.com, unless you have a local server you can use for exporting your git tree to the world? Then, people (esp. Jonathan / myself) can pull in your changes regularly and I'd expect 'em to appear in the next release. cheers. -- Nathan From kenj@internode.on.net Mon Nov 24 19:32:07 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=unavailable version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAP1W7nA029091 for ; Mon, 24 Nov 2008 19:32:07 -0600 X-ASG-Debug-ID: 1227576724-2f9602a90000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from mail.internode.on.net (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id BA08E1223BBC for ; Mon, 24 Nov 2008 17:32:05 -0800 (PST) Received: from mail.internode.on.net (bld-mail06.adl2.internode.on.net [203.16.214.70]) by cuda.sgi.com with ESMTP id cMBZhzNIwsGHVb0K for ; Mon, 24 Nov 2008 17:32:05 -0800 (PST) Received: from [192.168.1.100] (unverified [118.208.222.17]) by mail.internode.on.net (SurgeMail 3.8f2) with ESMTP id 34979715-1927428 for multiple; Tue, 25 Nov 2008 11:42:58 +1030 (CDT) X-ASG-Orig-Subj: Re: Daylight savings bug Subject: Re: Daylight savings bug From: Ken McDonell Reply-To: kenj@internode.on.net To: Nathan Scott Cc: pcp@oss.sgi.com In-Reply-To: <1227570519.4557.24.camel@verge.scott.net.au> References: <1223438632.4494.17.camel@verge.scott.net.au> <1227549185.10096.13.camel@bozo> <1227566299.4557.10.camel@verge.scott.net.au> <1227569919.10096.29.camel@bozo> <1227570519.4557.24.camel@verge.scott.net.au> Content-Type: text/plain Date: Tue, 25 Nov 2008 12:12:41 +1100 Message-Id: <1227575561.10096.47.camel@bozo> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: bld-mail06.adl2.internode.on.net[203.16.214.70] X-Barracuda-Start-Time: 1227576726 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -2.02 X-Barracuda-Spam-Status: No, SCORE=-2.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests= X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11076 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- Cleverly, we don't report the DAY, just the TIME in the pmdumplog output ... grr. But, looks like it _really_ is for Oct 6! $ date --date='TZ="UTC" 1970-01-01 00:00:00 +1223212334.242047sec' Mon Oct 6 00:12:14 EST 2008 My theory is not correct unfortunately. But I am sure pmdate reports the wrong thing in this boundary case, so on the basis that it should be fixed, and I don't have any more evidence from this particular event, and another one is not going to happen for at best 6 months. On Tue, 2008-11-25 at 10:48 +1100, Nathan Scott wrote: > On Tue, 2008-11-25 at 10:38 +1100, Ken McDonell wrote: > > Hmmm. > > > > Does not disprove my hypothesis, but does not prove it either. > > > > 20081006 is a bit odd. If pmdumplog -l reports an end time then the > > archive is not really busted. But pmdumplog -v is a strange beast, that > > looks like a pmdumplog bug (sigh). > > > > Just try pmdumplog 20081006 | head -100 ... the timestamps appear in the > > pmResult dumps. > > I've attached two dump heads - one pmdumplog with no options, > the other with -Dpdu (this one shows the raw timestamps). > > > But 20081006 looks like it is for the correct day which undermines my > > theory. > > > > I believe the problem is in pmdate -1d when you're within an hour of > > midnight and there was a DST adjustment in the previous 23-25 hrs ... > > there is no good answer here, pmdate cannot produce the 100% right > > answer no matter what. > > Yep, sounds entirely feasible. > > > I'm looking at reworking pmlogger_daily to remove all dependence on > > pmdate, and this will give me a chance to re-instate the -o option ... > > actually new behaviour to merge _all_ archives, not just yesterday's and > > -o for the old (current) behaviour, so all of pmlogger_daily QA will > > work again ... > > OK, great! > > cheers. > > -- > Nathan From markgw@sgi.com Mon Nov 24 19:38:13 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=FM_IS_IT_OUR_ACCOUNT autolearn=unavailable version=3.3.0-rupdated Received: from relay.sgi.com (relay3.corp.sgi.com [198.149.34.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAP1cCik029410 for ; Mon, 24 Nov 2008 19:38:13 -0600 Received: from larry.melbourne.sgi.com (larry.melbourne.sgi.com [134.14.52.130]) by relay3.corp.sgi.com (Postfix) with SMTP id A9C3DAC018; Mon, 24 Nov 2008 17:38:08 -0800 (PST) Received: from [134.14.55.22] (dhcp22.melbourne.sgi.com [134.14.55.22]) by larry.melbourne.sgi.com (950413.SGI.8.6.12/950213.SGI.AUTOCF) via ESMTP id MAA29211; Tue, 25 Nov 2008 12:38:03 +1100 Message-ID: <492B56FB.40604@sgi.com> Date: Tue, 25 Nov 2008 12:38:03 +1100 From: Mark Goodwin Reply-To: markgw@sgi.com Organization: SGI Engineering User-Agent: Thunderbird 2.0.0.18 (Windows/20081105) MIME-Version: 1.0 To: Nathan Scott CC: kenj@internode.on.net, pcp@oss.sgi.com Subject: Re: [pcp] Daylight savings bug - fix for pmdumplog -v References: <1223438632.4494.17.camel@verge.scott.net.au> <1227549185.10096.13.camel@bozo> <1227566299.4557.10.camel@verge.scott.net.au> <1227571952.10096.34.camel@bozo> <1227573397.4557.34.camel@verge.scott.net.au> In-Reply-To: <1227573397.4557.34.camel@verge.scott.net.au> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Nathan Scott wrote: > On Tue, 2008-11-25 at 11:12 +1100, Ken McDonell wrote: >> On Tue, 2008-11-25 at 09:38 +1100, Nathan Scott wrote: >>> ... >> I have a patch to fix this (missing endian conversion from the year >> dot). >> >> Patch is attached, which raises the question of the "correct" protocol >> for bozos outside the sgi-aconex consortium to get pcp patches into the >> one true official tree ... guidance? > > 1. Clone yerself a git tree from git://oss.sgi.com/pcp/pcp.git > 2. Create a branch for development > 3. Code, test, commit patches locally > 4. Push changes out to a visible server > 5. Send mail to pcp@oss.sgi.com with updates > > Probably best if you re-activate your account on oss.sgi.com, > unless you have a local server you can use for exporting your > git tree to the world? You can just post the patch to the list for review if you want, and if accepted, Jonathan will push it back out to the 'dev' branch at git://oss.sgi.com/pcp/pcp.git > Then, people (esp. Jonathan / myself) can pull in your changes > regularly and I'd expect 'em to appear in the next release. They'll appear in the next release (whenever that happens to be), but more expediently, in the dev branch as soon as Jonathan pushes back out to oss (which would be daily or on demand as needed). I can get oss accounts created if needed, but I'm wondering if we could maybe have a semi-open-access 'incoming' repository or something for those without a net visible server ..? Or will simply mailing patches to pcp@oss suffice? Cheers -- Mark From kenj@internode.on.net Mon Nov 24 21:02:54 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=unavailable version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAP32seN001802 for ; Mon, 24 Nov 2008 21:02:54 -0600 X-ASG-Debug-ID: 1227582171-3833015b0000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from mail.internode.on.net (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id A8BCF131226C for ; Mon, 24 Nov 2008 19:02:52 -0800 (PST) Received: from mail.internode.on.net (bld-mail11.adl2.internode.on.net [203.16.214.75]) by cuda.sgi.com with ESMTP id oYW89Z27ohhK93S6 for ; Mon, 24 Nov 2008 19:02:52 -0800 (PST) Received: from [192.168.1.100] (unverified [118.208.222.17]) by mail.internode.on.net (SurgeMail 3.8f2) with ESMTP id 36946958-1927428 for multiple; Tue, 25 Nov 2008 10:08:56 +1030 (CDT) X-ASG-Orig-Subj: Re: Daylight savings bug Subject: Re: Daylight savings bug From: Ken McDonell Reply-To: kenj@internode.on.net To: Nathan Scott Cc: pcp@oss.sgi.com In-Reply-To: <1227566299.4557.10.camel@verge.scott.net.au> References: <1223438632.4494.17.camel@verge.scott.net.au> <1227549185.10096.13.camel@bozo> <1227566299.4557.10.camel@verge.scott.net.au> Content-Type: text/plain Date: Tue, 25 Nov 2008 10:38:39 +1100 Message-Id: <1227569919.10096.29.camel@bozo> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: bld-mail11.adl2.internode.on.net[203.16.214.75] X-Barracuda-Start-Time: 1227582173 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -2.02 X-Barracuda-Spam-Status: No, SCORE=-2.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests= X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11081 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- Hmmm. Does not disprove my hypothesis, but does not prove it either. 20081006 is a bit odd. If pmdumplog -l reports an end time then the archive is not really busted. But pmdumplog -v is a strange beast, that looks like a pmdumplog bug (sigh). Just try pmdumplog 20081006 | head -100 ... the timestamps appear in the pmResult dumps. But 20081006 looks like it is for the correct day which undermines my theory. I believe the problem is in pmdate -1d when you're within an hour of midnight and there was a DST adjustment in the previous 23-25 hrs ... there is no good answer here, pmdate cannot produce the 100% right answer no matter what. I'm looking at reworking pmlogger_daily to remove all dependence on pmdate, and this will give me a chance to re-instate the -o option ... actually new behaviour to merge _all_ archives, not just yesterday's and -o for the old (current) behaviour, so all of pmlogger_daily QA will work again ... all of this was lost in one of the great rattle filesystem corruption disasters. On Tue, 2008-11-25 at 09:38 +1100, Nathan Scott wrote: > On Tue, 2008-11-25 at 04:53 +1100, Ken McDonell wrote: > > I may have a handle on this. > > > > Nathan, can you please check the following for me. > > > > 1. the summary archive for 20081004 is ok. > > Yep. > > > 2 the summary archive for 20081005 is missing. > > Yep. > > > 3. now can you check the summary archive for 20081006 ... I'm interested > > in the actual dates for the label and the log records in that > > archive ... I suspect it is really data for 20081005. > > See below, appears to really be data for 06 AFAICT. > > > 4. and if that is the case, I suspect you also have the component > > archives for 20081006 that have not been merged (although these may well > > have been culled by now). > > Yeah, no component archives left anywhere now that I can see. > > $ pmdumplog -l 20081004 > Log Label (Log Format Version 2) > Performance metrics from host app4 > commencing Sat Oct 4 00:13:41.741 2008 > ending Sun Oct 5 00:13:26.753 2008 > $ pmdumplog -l 20081006 > Log Label (Log Format Version 2) > Performance metrics from host app4 > commencing Mon Oct 6 00:12:14.239 2008 > ending Tue Oct 7 00:13:35.340 2008 > > All options I tried to pmdumplog give timestamps relative > to start of the day (and all appeared to be from the start > of 06). Trying to get more info with -v was unhelpful - a > bug lurking here too perhaps? All archives produce this: > > $ pmdumplog -v 20081006.0 > Raw dump of physical archive file "20081006.0" ... > Dump ... record len: -2080374784 @ offset: 0 > Trailer botch: 36046160 != -2080374784 > > Is there a better way to dump out the timestamps than with > pmdumplog? (or an option I've missed? -Dsomething?) > > thanks! > > -- > Nathan > From kenj@internode.on.net Mon Nov 24 21:08:27 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=FM_IS_IT_OUR_ACCOUNT autolearn=unavailable version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAP38RY1002056 for ; Mon, 24 Nov 2008 21:08:27 -0600 X-ASG-Debug-ID: 1227582505-3d9a013d0000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from mail.internode.on.net (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 48A2815E4229 for ; Mon, 24 Nov 2008 19:08:26 -0800 (PST) Received: from mail.internode.on.net (bld-mail06.adl2.internode.on.net [203.16.214.70]) by cuda.sgi.com with ESMTP id B0qHDHTIIaZ85AGr for ; Mon, 24 Nov 2008 19:08:26 -0800 (PST) Received: from [192.168.1.100] (unverified [118.208.222.17]) by mail.internode.on.net (SurgeMail 3.8f2) with ESMTP id 34991413-1927428 for multiple; Tue, 25 Nov 2008 13:35:12 +1030 (CDT) X-ASG-Orig-Subj: Re: [pcp] Daylight savings bug - fix for pmdumplog -v Subject: Re: [pcp] Daylight savings bug - fix for pmdumplog -v From: Ken McDonell Reply-To: kenj@internode.on.net To: Nathan Scott Cc: markgw@sgi.com, pcp@oss.sgi.com In-Reply-To: <1227581292.4557.45.camel@verge.scott.net.au> References: <1223438632.4494.17.camel@verge.scott.net.au> <1227549185.10096.13.camel@bozo> <1227566299.4557.10.camel@verge.scott.net.au> <1227571952.10096.34.camel@bozo> <1227573397.4557.34.camel@verge.scott.net.au> <492B56FB.40604@sgi.com> <1227581292.4557.45.camel@verge.scott.net.au> Content-Type: text/plain Date: Tue, 25 Nov 2008 14:04:57 +1100 Message-Id: <1227582297.10096.55.camel@bozo> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: bld-mail06.adl2.internode.on.net[203.16.214.70] X-Barracuda-Start-Time: 1227582507 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0002 1.0000 -2.0196 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -2.02 X-Barracuda-Spam-Status: No, SCORE=-2.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests= X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11081 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- I don't mind either way. If I'm going the git path then I will need Mark to get a login for me on oss.sgi.com. For all sorts of reasons this would be easier for me if it was "kenj" rather than the old sgi "kenmcd" login. Mark, over to you. On Tue, 2008-11-25 at 13:48 +1100, Nathan Scott wrote: > On Tue, 2008-11-25 at 12:38 +1100, Mark Goodwin wrote: > > > On Tue, 2008-11-25 at 11:12 +1100, Ken McDonell wrote: > > >> On Tue, 2008-11-25 at 09:38 +1100, Nathan Scott wrote: > > > Probably best if you re-activate your account on oss.sgi.com, > > > unless you have a local server you can use for exporting your > > > git tree to the world? > > > > You can just post the patch to the list for review if you want, and > > if accepted, Jonathan will push it back out to the 'dev' branch at > > git://oss.sgi.com/pcp/pcp.git > > For someone working regularly, this model ends up pushing > alot of extra work onto other individuals, instead of > spreading the load - this is happening already with just > Jonathan & I regularly doing git work (eg. domain numbers, > pcpweb updates, etc) ... it would be better for someone > with Kens background to be using git too IMO. > > > I can get oss accounts created if needed, but I'm wondering if > > we could maybe have a semi-open-access 'incoming' repository or > > something for those without a net visible server ..? Or will simply > > mailing patches to pcp@oss suffice? > > For most people I think that'd be fine, but for someone > with lots of history & high levels of trust from everyone > working on PCP already, it'd be better to go the git route. > > cheers. > > -- > Nathan > From nscott@aconex.com Mon Nov 24 21:08:48 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=FM_IS_IT_OUR_ACCOUNT autolearn=unavailable version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAP38mJ0002107 for ; Mon, 24 Nov 2008 21:08:48 -0600 X-ASG-Debug-ID: 1227582527-3d7901410000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from postoffice2.aconex.com (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 82D8A15E4AA8 for ; Mon, 24 Nov 2008 19:08:47 -0800 (PST) Received: from postoffice2.aconex.com (mail.aconex.com [203.89.202.182]) by cuda.sgi.com with ESMTP id ELVUWLrD8ZWIAk3K for ; Mon, 24 Nov 2008 19:08:47 -0800 (PST) Received: from postoffice.aconex.com (localhost [127.0.0.1]) by postoffice2.aconex.com (Spam Firewall) with ESMTP id C553940385D; Tue, 25 Nov 2008 13:50:31 +1100 (EST) Received: from postoffice.aconex.com (postoffice.yarra.acx [192.168.102.1]) by postoffice2.aconex.com with ESMTP id 9n5rya0zxvdWiDi4; Tue, 25 Nov 2008 13:50:31 +1100 (EST) Received: from [192.168.5.24] (melho0.aconex.com [203.89.192.141]) by postoffice.aconex.com (Postfix) with ESMTP id 855E892C2E2; Tue, 25 Nov 2008 13:50:31 +1100 (EST) X-ASG-Orig-Subj: Re: [pcp] Daylight savings bug - fix for pmdumplog -v Subject: Re: [pcp] Daylight savings bug - fix for pmdumplog -v From: Nathan Scott To: markgw@sgi.com Cc: kenj@internode.on.net, pcp@oss.sgi.com In-Reply-To: <492B56FB.40604@sgi.com> References: <1223438632.4494.17.camel@verge.scott.net.au> <1227549185.10096.13.camel@bozo> <1227566299.4557.10.camel@verge.scott.net.au> <1227571952.10096.34.camel@bozo> <1227573397.4557.34.camel@verge.scott.net.au> <492B56FB.40604@sgi.com> Content-Type: text/plain Date: Tue, 25 Nov 2008 13:48:12 +1100 Message-Id: <1227581292.4557.45.camel@verge.scott.net.au> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: mail.aconex.com[203.89.202.182] X-Barracuda-Start-Time: 1227582528 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0208 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -2.02 X-Barracuda-Spam-Status: No, SCORE=-2.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests= X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11081 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- On Tue, 2008-11-25 at 12:38 +1100, Mark Goodwin wrote: > > On Tue, 2008-11-25 at 11:12 +1100, Ken McDonell wrote: > >> On Tue, 2008-11-25 at 09:38 +1100, Nathan Scott wrote: > > Probably best if you re-activate your account on oss.sgi.com, > > unless you have a local server you can use for exporting your > > git tree to the world? > > You can just post the patch to the list for review if you want, and > if accepted, Jonathan will push it back out to the 'dev' branch at > git://oss.sgi.com/pcp/pcp.git For someone working regularly, this model ends up pushing alot of extra work onto other individuals, instead of spreading the load - this is happening already with just Jonathan & I regularly doing git work (eg. domain numbers, pcpweb updates, etc) ... it would be better for someone with Kens background to be using git too IMO. > I can get oss accounts created if needed, but I'm wondering if > we could maybe have a semi-open-access 'incoming' repository or > something for those without a net visible server ..? Or will simply > mailing patches to pcp@oss suffice? For most people I think that'd be fine, but for someone with lots of history & high levels of trust from everyone working on PCP already, it'd be better to go the git route. cheers. -- Nathan From kenj@internode.on.net Mon Nov 24 21:47:39 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=FM_IS_IT_OUR_ACCOUNT autolearn=unavailable version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAP3lc7J004316 for ; Mon, 24 Nov 2008 21:47:39 -0600 X-ASG-Debug-ID: 1227584854-391a02d90000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from mail.internode.on.net (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 9704B15E723A for ; Mon, 24 Nov 2008 19:47:34 -0800 (PST) Received: from mail.internode.on.net (bld-mail09.adl2.internode.on.net [203.16.214.73]) by cuda.sgi.com with ESMTP id kfRgdTKLq1q2c1z9 for ; Mon, 24 Nov 2008 19:47:34 -0800 (PST) Received: from [192.168.1.100] (unverified [118.208.222.17]) by mail.internode.on.net (SurgeMail 3.8f2) with ESMTP id 49258746-1927428 for multiple; Tue, 25 Nov 2008 12:54:53 +1030 (CDT) X-ASG-Orig-Subj: Re: [pcp] Daylight savings bug - fix for pmdumplog -v Subject: Re: [pcp] Daylight savings bug - fix for pmdumplog -v From: Ken McDonell Reply-To: kenj@internode.on.net To: markgw@sgi.com Cc: Nathan Scott , pcp@oss.sgi.com In-Reply-To: <492B56FB.40604@sgi.com> References: <1223438632.4494.17.camel@verge.scott.net.au> <1227549185.10096.13.camel@bozo> <1227566299.4557.10.camel@verge.scott.net.au> <1227571952.10096.34.camel@bozo> <1227573397.4557.34.camel@verge.scott.net.au> <492B56FB.40604@sgi.com> Content-Type: text/plain Date: Tue, 25 Nov 2008 13:24:36 +1100 Message-Id: <1227579876.10096.52.camel@bozo> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: bld-mail09.adl2.internode.on.net[203.16.214.73] X-Barracuda-Start-Time: 1227584857 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0001 1.0000 -2.0207 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -2.02 X-Barracuda-Spam-Status: No, SCORE=-2.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests= X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11083 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- Mark, Provided someone acts as the gatekeeper and patches don't disappear into the bog of eternal stench, mailing 'em to the list works just fine for me. On Tue, 2008-11-25 at 12:38 +1100, Mark Goodwin wrote: > > Nathan Scott wrote: > > On Tue, 2008-11-25 at 11:12 +1100, Ken McDonell wrote: > >> On Tue, 2008-11-25 at 09:38 +1100, Nathan Scott wrote: > >>> ... > >> I have a patch to fix this (missing endian conversion from the year > >> dot). > >> > >> Patch is attached, which raises the question of the "correct" protocol > >> for bozos outside the sgi-aconex consortium to get pcp patches into the > >> one true official tree ... guidance? > > > > 1. Clone yerself a git tree from git://oss.sgi.com/pcp/pcp.git > > 2. Create a branch for development > > 3. Code, test, commit patches locally > > 4. Push changes out to a visible server > > 5. Send mail to pcp@oss.sgi.com with updates > > > > Probably best if you re-activate your account on oss.sgi.com, > > unless you have a local server you can use for exporting your > > git tree to the world? > > You can just post the patch to the list for review if you want, and > if accepted, Jonathan will push it back out to the 'dev' branch at > git://oss.sgi.com/pcp/pcp.git > > > Then, people (esp. Jonathan / myself) can pull in your changes > > regularly and I'd expect 'em to appear in the next release. > > They'll appear in the next release (whenever that happens to be), > but more expediently, in the dev branch as soon as Jonathan pushes > back out to oss (which would be daily or on demand as needed). > > I can get oss accounts created if needed, but I'm wondering if > we could maybe have a semi-open-access 'incoming' repository or > something for those without a net visible server ..? Or will simply > mailing patches to pcp@oss suffice? > > Cheers > -- Mark From kenj@internode.on.net Mon Nov 24 23:31:42 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.2 required=5.0 tests=J_CHICKENPOX_51, J_CHICKENPOX_74 autolearn=no version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAP5Ve1T010079 for ; Mon, 24 Nov 2008 23:31:42 -0600 X-ASG-Debug-ID: 1227591097-1093033a0000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from mail.internode.on.net (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 11D8C15E75AF for ; Mon, 24 Nov 2008 21:31:38 -0800 (PST) Received: from mail.internode.on.net (bld-mail09.adl2.internode.on.net [203.16.214.73]) by cuda.sgi.com with ESMTP id X7gIxR8opVaOvtjo for ; Mon, 24 Nov 2008 21:31:38 -0800 (PST) Received: from [192.168.1.100] (unverified [118.208.222.17]) by mail.internode.on.net (SurgeMail 3.8f2) with ESMTP id 49246045-1927428 for multiple; Tue, 25 Nov 2008 10:42:48 +1030 (CDT) X-ASG-Orig-Subj: Re: Daylight savings bug - fix for pmdumplog -v Subject: Re: Daylight savings bug - fix for pmdumplog -v From: Ken McDonell Reply-To: kenj@internode.on.net To: Nathan Scott Cc: pcp@oss.sgi.com In-Reply-To: <1227566299.4557.10.camel@verge.scott.net.au> References: <1223438632.4494.17.camel@verge.scott.net.au> <1227549185.10096.13.camel@bozo> <1227566299.4557.10.camel@verge.scott.net.au> Content-Type: multipart/mixed; boundary="=-1EbL9SH92sFlzIfOz9l/" Date: Tue, 25 Nov 2008 11:12:32 +1100 Message-Id: <1227571952.10096.34.camel@bozo> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 X-Barracuda-Connect: bld-mail09.adl2.internode.on.net[203.16.214.73] X-Barracuda-Start-Time: 1227591100 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -2.02 X-Barracuda-Spam-Status: No, SCORE=-2.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests= X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11089 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- --=-1EbL9SH92sFlzIfOz9l/ Content-Type: text/plain Content-Transfer-Encoding: 7bit On Tue, 2008-11-25 at 09:38 +1100, Nathan Scott wrote: > ... > > All options I tried to pmdumplog give timestamps relative > to start of the day (and all appeared to be from the start > of 06). Trying to get more info with -v was unhelpful - a > bug lurking here too perhaps? All archives produce this: > > $ pmdumplog -v 20081006.0 > Raw dump of physical archive file "20081006.0" ... > Dump ... record len: -2080374784 @ offset: 0 > Trailer botch: 36046160 != -2080374784 > > Is there a better way to dump out the timestamps than with > pmdumplog? (or an option I've missed? -Dsomething?) I have a patch to fix this (missing endian conversion from the year dot). Patch is attached, which raises the question of the "correct" protocol for bozos outside the sgi-aconex consortium to get pcp patches into the one true official tree ... guidance? --=-1EbL9SH92sFlzIfOz9l/ Content-Disposition: attachment; filename=pmdumplog-v.patch Content-Type: text/x-patch; name=pmdumplog-v.patch; charset=UTF-8 Content-Transfer-Encoding: 7bit diff --git a/src/pmdumplog/pmdumplog.c b/src/pmdumplog/pmdumplog.c index 9ed4358..6f17ee1 100644 --- a/src/pmdumplog/pmdumplog.c +++ b/src/pmdumplog/pmdumplog.c @@ -293,9 +293,10 @@ rawdump(FILE *f) int sts; old = ftell(f); - fseek(f, 0, SEEK_SET); + fseek(f, (long)0, SEEK_SET); while ((sts = fread(&len, 1, sizeof(len), f)) == sizeof(len)) { + len = ntohl(len); printf("Dump ... record len: %d @ offset: %ld", len, ftell(f) - sizeof(len)); len -= 2 * sizeof(len); for (i = 0; i < len; i++) { @@ -306,7 +307,7 @@ rawdump(FILE *f) } if (i % 32 == 0) putchar('\n'); if (i % 4 == 0) putchar(' '); - printf("%02x", check &0xff); + printf("%02x", check & 0xff); } putchar('\n'); if ((sts = fread(&check, 1, sizeof(check), f)) != sizeof(check)) { @@ -314,6 +315,7 @@ rawdump(FILE *f) printf("Unexpected EOF\n"); break; } + check = ntohl(check); len += 2 * sizeof(len); if (check != len) { printf("Trailer botch: %d != %d\n", check, len); --=-1EbL9SH92sFlzIfOz9l/-- From nscott@aconex.com Tue Nov 25 00:00:30 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=unavailable version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAP60Up8013755 for ; Tue, 25 Nov 2008 00:00:30 -0600 X-ASG-Debug-ID: 1227592826-35d4027a0000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from postoffice2.aconex.com (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 935DC15E787D for ; Mon, 24 Nov 2008 22:00:26 -0800 (PST) Received: from postoffice2.aconex.com (mail.aconex.com [203.89.202.182]) by cuda.sgi.com with ESMTP id kyJZ5hNX57QSiACW for ; Mon, 24 Nov 2008 22:00:26 -0800 (PST) Received: from postoffice.aconex.com (localhost [127.0.0.1]) by postoffice2.aconex.com (Spam Firewall) with ESMTP id 76D6340AC08 for ; Tue, 25 Nov 2008 16:59:25 +1100 (EST) Received: from postoffice.aconex.com (postoffice.yarra.acx [192.168.102.1]) by postoffice2.aconex.com with ESMTP id UL8Rb0HEXkWhsnx6 for ; Tue, 25 Nov 2008 16:59:25 +1100 (EST) Received: from [192.168.5.24] (melho0.aconex.com [203.89.192.141]) by postoffice.aconex.com (Postfix) with ESMTP id 582FE92C2F9 for ; Tue, 25 Nov 2008 16:59:25 +1100 (EST) X-ASG-Orig-Subj: pcp updates Subject: pcp updates From: Nathan Scott To: pcp@oss.sgi.com Content-Type: text/plain Date: Tue, 25 Nov 2008 16:57:06 +1100 Message-Id: <1227592626.4557.52.camel@verge.scott.net.au> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: mail.aconex.com[203.89.202.182] X-Barracuda-Start-Time: 1227592829 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -1.52 X-Barracuda-Spam-Status: No, SCORE=-1.52 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests=BSF_RULE7568M X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11091 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- 0.50 BSF_RULE7568M Custom Rule 7568M Changes committed to git://oss.sgi.com/nathans/pcp.git src/pmdas/windows/README | 2 src/pmdas/windows/hypnotoad.h | 5 src/pmdas/windows/instance.c | 30 ++-- src/pmdas/windows/libpdh.def | 1 src/pmdas/windows/libpdh.h | 3 src/pmdas/windows/open.c | 304 ++++++++++++++++++++---------------------- src/pmdas/windows/pmda.c | 5 src/pmdumplog/pmdumplog.c | 6 8 files changed, 182 insertions(+), 174 deletions(-) commit 0469b3aa229d83c2d08c608807bf0deb0810e77d Author: Ken McDonell Date: Tue Nov 25 16:53:16 2008 +1100 Add a missing endian conversion in pmdumplog "raw" mode, as reported here... On Tue, 2008-11-25 at 09:38 +1100, Nathan Scott wrote: > ... . Trying to get more info with -v was unhelpful - a > bug lurking here too perhaps? All archives produce this: > > $ pmdumplog -v 20081006.0 > Raw dump of physical archive file "20081006.0" ... > Dump ... record len: -2080374784 @ offset: 0 > Trailer botch: 36046160 != -2080374784 commit a552bc84f0ca5ff1a63b86a0183daa884ca4c802 Author: Nathan Scott Date: Tue Nov 25 16:43:54 2008 +1100 Switch to using PdhRemoveCounter as recommended by the Windows API documentation. commit 2bd9f0b05441fe0466da6bd9f0b384b24bb07fcf Author: Nathan Scott Date: Tue Nov 25 16:40:22 2008 +1100 Windows PMDA fixes. Fix some of the pmdaCache uses that were not passing in the indom identifier with the domain number OR'd in. Fix the M_REDO flag, so that instances which require full refresh are in fact refreshed on instance lookups. Finally, the network interface instance names now use the disk indom trick of replacing spaces with underscores to ensure the PCP instance domain rules are followed. commit bfab9527dddd7c3fbefa761706f47b70ff94bbb7 Author: Nathan Scott Date: Tue Nov 25 16:38:53 2008 +1100 Fix typo in Windows PMDA README file. From nscott@aconex.com Tue Nov 25 00:04:27 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=unavailable version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAP64R1o014362 for ; Tue, 25 Nov 2008 00:04:27 -0600 X-ASG-Debug-ID: 1227593065-35d503390000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from postoffice2.aconex.com (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id E6FA815E78AF for ; Mon, 24 Nov 2008 22:04:26 -0800 (PST) Received: from postoffice2.aconex.com (mail.aconex.com [203.89.202.182]) by cuda.sgi.com with ESMTP id OmJGPECXZbH7SxlE for ; Mon, 24 Nov 2008 22:04:26 -0800 (PST) Received: from postoffice.aconex.com (localhost [127.0.0.1]) by postoffice2.aconex.com (Spam Firewall) with ESMTP id 0F41440B6AC; Tue, 25 Nov 2008 17:04:25 +1100 (EST) Received: from postoffice.aconex.com (postoffice.yarra.acx [192.168.102.1]) by postoffice2.aconex.com with ESMTP id YibXeG1a5GYht6cN; Tue, 25 Nov 2008 17:04:25 +1100 (EST) Received: from [192.168.5.24] (melho0.aconex.com [203.89.192.141]) by postoffice.aconex.com (Postfix) with ESMTP id DACA592C2E2; Tue, 25 Nov 2008 17:04:24 +1100 (EST) X-ASG-Orig-Subj: Re: [pcp] Daylight savings bug - fix for pmdumplog -v Subject: Re: [pcp] Daylight savings bug - fix for pmdumplog -v From: Nathan Scott To: kenj@internode.on.net Cc: markgw@sgi.com, pcp@oss.sgi.com In-Reply-To: <1227582297.10096.55.camel@bozo> References: <1223438632.4494.17.camel@verge.scott.net.au> <1227549185.10096.13.camel@bozo> <1227566299.4557.10.camel@verge.scott.net.au> <1227571952.10096.34.camel@bozo> <1227573397.4557.34.camel@verge.scott.net.au> <492B56FB.40604@sgi.com> <1227581292.4557.45.camel@verge.scott.net.au> <1227582297.10096.55.camel@bozo> Content-Type: text/plain Date: Tue, 25 Nov 2008 17:02:06 +1100 Message-Id: <1227592926.4557.58.camel@verge.scott.net.au> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: mail.aconex.com[203.89.202.182] X-Barracuda-Start-Time: 1227593066 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -2.02 X-Barracuda-Spam-Status: No, SCORE=-2.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests= X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11091 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- On Tue, 2008-11-25 at 14:04 +1100, Ken McDonell wrote: > I don't mind either way. > > If I'm going the git path then I will need Mark to get a login for me on > oss.sgi.com. > ... I've tested the fix & merged it in my tree since I was pushing out some Windows changes anyway. With any luck, the next one will come in via a git-pull! cheers. -- Nathan From markgw@sgi.com Tue Nov 25 00:29:46 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=unavailable version=3.3.0-rupdated Received: from relay.sgi.com (relay2.corp.sgi.com [137.38.102.29]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAP6Tkx6015538 for ; Tue, 25 Nov 2008 00:29:46 -0600 Received: from larry.melbourne.sgi.com (larry.melbourne.sgi.com [134.14.52.130]) by relay2.corp.sgi.com (Postfix) with SMTP id 9A7513040A6; Mon, 24 Nov 2008 22:29:41 -0800 (PST) Received: from [134.14.55.22] (dhcp22.melbourne.sgi.com [134.14.55.22]) by larry.melbourne.sgi.com (950413.SGI.8.6.12/950213.SGI.AUTOCF) via ESMTP id RAA07199; Tue, 25 Nov 2008 17:29:36 +1100 Message-ID: <492B9B50.7090004@sgi.com> Date: Tue, 25 Nov 2008 17:29:36 +1100 From: Mark Goodwin Reply-To: markgw@sgi.com Organization: SGI Engineering User-Agent: Thunderbird 2.0.0.18 (Windows/20081105) MIME-Version: 1.0 To: Nathan Scott CC: kenj@internode.on.net, pcp@oss.sgi.com Subject: Re: [pcp] Daylight savings bug - fix for pmdumplog -v References: <1223438632.4494.17.camel@verge.scott.net.au> <1227549185.10096.13.camel@bozo> <1227566299.4557.10.camel@verge.scott.net.au> <1227571952.10096.34.camel@bozo> <1227573397.4557.34.camel@verge.scott.net.au> <492B56FB.40604@sgi.com> <1227581292.4557.45.camel@verge.scott.net.au> <1227582297.10096.55.camel@bozo> <1227592926.4557.58.camel@verge.scott.net.au> In-Reply-To: <1227592926.4557.58.camel@verge.scott.net.au> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Nathan Scott wrote: > On Tue, 2008-11-25 at 14:04 +1100, Ken McDonell wrote: >> I don't mind either way. >> >> If I'm going the git path then I will need Mark to get a login for me on >> oss.sgi.com. >> ... > > I've tested the fix & merged it in my tree since I was pushing > out some Windows changes anyway. and Jonathan will pull it in via Nathan's tree. > With any luck, the next one will come in via a git-pull! Not sure on the delay with account creation due to USA thanksgiving. Cheers -- Mark ps: the diplomatic term for Bog of Eternal Stench is 'quagmire', but fear not, git is our Saviour :) From nscott@aconex.com Tue Nov 25 02:48:48 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=unavailable version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAP8mmLn024407 for ; Tue, 25 Nov 2008 02:48:48 -0600 X-ASG-Debug-ID: 1227602926-445a02ef0000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from postoffice2.aconex.com (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id C63021BE15F2 for ; Tue, 25 Nov 2008 00:48:46 -0800 (PST) Received: from postoffice2.aconex.com (mail.aconex.com [203.89.202.182]) by cuda.sgi.com with ESMTP id 1O21oazhnrXHEYbu for ; Tue, 25 Nov 2008 00:48:46 -0800 (PST) Received: from postoffice.aconex.com (localhost [127.0.0.1]) by postoffice2.aconex.com (Spam Firewall) with ESMTP id CDCF83FE44A for ; Mon, 24 Nov 2008 21:55:16 +1100 (EST) Received: from postoffice.aconex.com (postoffice.yarra.acx [192.168.102.1]) by postoffice2.aconex.com with ESMTP id MBe9Eos0UUqqeBNC for ; Mon, 24 Nov 2008 21:55:16 +1100 (EST) Received: from [192.168.0.100] (c220-239-214-222.fernt2.vic.optusnet.com.au [220.239.214.222]) by postoffice.aconex.com (Postfix) with ESMTP id B32E792C2F9 for ; Mon, 24 Nov 2008 21:55:16 +1100 (EST) X-ASG-Orig-Subj: pcp updates Subject: pcp updates From: Nathan Scott To: pcp@oss.sgi.com Content-Type: text/plain Date: Mon, 24 Nov 2008 21:52:58 +1100 Message-Id: <1227523978.5749.7.camel@verge.scott.net.au> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: mail.aconex.com[203.89.202.182] X-Barracuda-Start-Time: 1227602927 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -1.52 X-Barracuda-Spam-Status: No, SCORE=-1.52 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests=BSF_RULE7568M X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11101 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- 0.50 BSF_RULE7568M Custom Rule 7568M Changes committed to git://oss.sgi.com/nathans/pcp.git src/pmdas/dbping/README | 51 ------------- src/pmdas/dbping/dbprobe.pl | 16 ++-- src/pmdas/dbping/pmdadbping.pl | 83 +++++++++++++++++++-- src/pmdas/kvm/pmdakvm.pl | 12 +-- src/pmdas/memcache/README | 52 ------------- src/pmdas/memcache/pmdamemcache.pl | 137 +++++++++++++++++++++++++---------- src/pmdas/systemtap/pmdasystemtap.pl | 80 ++++++++++++++++++++ src/pmdas/vmware/pmdavmware.pl | 77 +++++++++++++++++++ src/pmns/stdpmid.add-on | 2 src/pmns/stdpmid.pcp | 2 10 files changed, 351 insertions(+), 161 deletions(-) commit 7552b66acae654c9e1fe067622d7870dc6b536e7 Author: Nathan Scott Date: Mon Nov 24 21:51:41 2008 +1100 Update the standard domain numbers to avoid conflicts once more. Shift the Lustre PMDA to slot 93, to match the SGI tree and avoid the memcache PMDA. Also shunt the KVM PMDA down to slot 95, since it'd inadvertantly been added in at 93 as well. What a zoo! :) commit 1a10e27ed8d33c60012a2b0dab22d3c680de0385 Author: Nathan Scott Date: Mon Nov 24 10:03:17 2008 +1100 Manual pages for many of the Perl PMDAs. We should be able to invoke pod2man to generate those for installation, but thats a job for another day (as are the last couple of remaining Perl PMDAs needing better docs). From nscott@aconex.com Tue Nov 25 16:28:49 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.1 required=5.0 tests=J_CHICKENPOX_48 autolearn=no version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAPMSnFD012952 for ; Tue, 25 Nov 2008 16:28:49 -0600 X-ASG-Debug-ID: 1227652127-57ba00c30000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from postoffice2.aconex.com (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 91A1715F552E for ; Tue, 25 Nov 2008 14:28:47 -0800 (PST) Received: from postoffice2.aconex.com (mail.aconex.com [203.89.202.182]) by cuda.sgi.com with ESMTP id Zp6JV18AHlepY6vk for ; Tue, 25 Nov 2008 14:28:47 -0800 (PST) Received: from postoffice.aconex.com (localhost [127.0.0.1]) by postoffice2.aconex.com (Spam Firewall) with ESMTP id 0D90C415DBA for ; Wed, 26 Nov 2008 08:50:17 +1100 (EST) Received: from postoffice.aconex.com (postoffice.yarra.acx [192.168.102.1]) by postoffice2.aconex.com with ESMTP id xMKEN1xWyMSJzlSx for ; Wed, 26 Nov 2008 08:50:17 +1100 (EST) Received: from [192.168.0.100] (c220-239-214-222.fernt2.vic.optusnet.com.au [220.239.214.222]) by postoffice.aconex.com (Postfix) with ESMTP id D9BCB92C2E2 for ; Wed, 26 Nov 2008 08:50:16 +1100 (EST) X-ASG-Orig-Subj: pcp updates Subject: pcp updates From: Nathan Scott To: pcp@oss.sgi.com Content-Type: text/plain Date: Wed, 26 Nov 2008 08:47:58 +1100 Message-Id: <1227649678.4399.40.camel@verge.scott.net.au> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: mail.aconex.com[203.89.202.182] X-Barracuda-Start-Time: 1227652128 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -2.02 X-Barracuda-Spam-Status: No, SCORE=-2.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests= X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11149 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- Changes committed to git://oss.sgi.com/nathans/pcp.git src/pmdas/memcache/GNUmakefile | 2 +- src/pmdas/windows/pmda.c | 32 +++++++++++++++++++++++++------- src/pmdas/windows/pmns.hinv | 2 ++ 3 files changed, 28 insertions(+), 8 deletions(-) commit 87bf1162aef138e7d9eec142f2cc446782511209 Author: Jonathan Knispel Date: Wed Nov 26 08:46:49 2008 +1100 Remove README from memcache PMDA, this info is in the embedded Perl POD within the pmdamemcache.pl script now. commit ecc2a160303d414ea4a4e454e6606cdf42ea5863 Author: Nathan Scott Date: Tue Nov 25 18:12:25 2008 +1100 Add the hinv.pagesize and hinv.nfilesys metrics into pmdawindows. From nscott@aconex.com Tue Nov 25 19:57:33 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=unavailable version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAQ1vWGh027451 for ; Tue, 25 Nov 2008 19:57:33 -0600 X-ASG-Debug-ID: 1227664651-633401fc0000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from postoffice2.aconex.com (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 67BAF1B8B54C for ; Tue, 25 Nov 2008 17:57:31 -0800 (PST) Received: from postoffice2.aconex.com (mail.aconex.com [203.89.202.182]) by cuda.sgi.com with ESMTP id bMtXBXOX5GNm5rPW for ; Tue, 25 Nov 2008 17:57:31 -0800 (PST) Received: from postoffice.aconex.com (localhost [127.0.0.1]) by postoffice2.aconex.com (Spam Firewall) with ESMTP id 23BEA4180A4 for ; Wed, 26 Nov 2008 12:56:30 +1100 (EST) Received: from postoffice.aconex.com (postoffice.yarra.acx [192.168.102.1]) by postoffice2.aconex.com with ESMTP id HC3Wy993q8K0g9VX for ; Wed, 26 Nov 2008 12:56:30 +1100 (EST) Received: from [192.168.0.100] (c220-239-214-222.fernt2.vic.optusnet.com.au [220.239.214.222]) by postoffice.aconex.com (Postfix) with ESMTP id EEF3F92C0E6 for ; Wed, 26 Nov 2008 12:56:29 +1100 (EST) X-ASG-Orig-Subj: fix "make clean" Subject: fix "make clean" From: Nathan Scott To: pcp@oss.sgi.com Content-Type: text/plain Date: Wed, 26 Nov 2008 12:54:10 +1100 Message-Id: <1227664450.4399.57.camel@verge.scott.net.au> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: mail.aconex.com[203.89.202.182] X-Barracuda-Start-Time: 1227664652 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -2.02 X-Barracuda-Spam-Status: No, SCORE=-2.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests= X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11163 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- Changes committed to git://oss.sgi.com/nathans/pcp.git src/include/GNUmakefile | 2 +- src/include/buildrules | 5 ++++- 2 files changed, 5 insertions(+), 2 deletions(-) commit 3581a649887c7a38f3d4b92e666e38bc814d4650 Author: Nathan Scott Date: Wed Nov 26 12:41:16 2008 +1100 Fix "make; make clean; make" which failed in the include directory because the LDIRT there contained half of the configure-generated files. This makes us deal the same way with all configure-generated files, and: - make clobber/distclean => completely clean - make clean" => clean such that a rebuild will not need to run the configure script again ... which was the intention with these different targets. From jkwaoz@sgi.com Wed Nov 26 00:04:05 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: ** X-Spam-Status: No, score=2.2 required=5.0 tests=J_CHICKENPOX_48, J_CHICKENPOX_73,URIBL_BLACK autolearn=no version=3.3.0-rupdated Received: from relay.sgi.com (relay1.corp.sgi.com [137.38.102.111]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAQ645aT017421 for ; Wed, 26 Nov 2008 00:04:05 -0600 Received: from [134.15.251.1] (melb-sw-corp-251-1.corp.sgi.com [134.15.251.1]) by relay1.corp.sgi.com (Postfix) with ESMTP id 642518F807F for ; Tue, 25 Nov 2008 22:04:00 -0800 (PST) Message-ID: <492CE6CE.1000909@sgi.com> Date: Wed, 26 Nov 2008 15:03:58 +0900 From: Jonathan Knispel User-Agent: Thunderbird 2.0.0.17 (X11/20080922) MIME-Version: 1.0 To: pcp@oss.sgi.com Subject: pcp updates: Performance Co-Pilot (PCP) branch, dev, updated. 2.7.7-20080924-48-gc3223ff X-Enigmail-Version: 0.95.7 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Commits from Nathan Scott's branch have been merged into the dev branch of the PCP git repository on oss.sgi.com. The details are below. [A technical hitch prevented this email being generated automatically, so this is a manual repost, with light editing, of our internal push email.] Regards, Jonathan ======== Updates for the "Performance Co-Pilot (PCP)" project have been committed to the repository at git://oss.sgi.com:8090/pcp/pcp.git The branch "dev" has been updated via c3223ffdfe4f8f5fe24d63aff21f17b61dbc57d5 (commit) via 1cf1f1609b11cb6a729b8d92ad882489a63d8b5c (commit) via 3581a649887c7a38f3d4b92e666e38bc814d4650 (commit) via 87bf1162aef138e7d9eec142f2cc446782511209 (commit) via ecc2a160303d414ea4a4e454e6606cdf42ea5863 (commit) via 0469b3aa229d83c2d08c608807bf0deb0810e77d (commit) via a552bc84f0ca5ff1a63b86a0183daa884ca4c802 (commit) via 2bd9f0b05441fe0466da6bd9f0b384b24bb07fcf (commit) via bfab9527dddd7c3fbefa761706f47b70ff94bbb7 (commit) via e0ed285f0b6d23878a349b493482aa5fbf682434 (commit) via 6ee1ccbf34b3e1e83c00374d4801e7c466efc780 (commit) via 7552b66acae654c9e1fe067622d7870dc6b536e7 (commit) via 1a10e27ed8d33c60012a2b0dab22d3c680de0385 (commit) via 80245342a195ae2cad607c84e564cef76fe0fbc4 (commit) via 32f96056e7f4e67e94da21f4b5a61f7d05bb749d (commit) via beb84d862142fc5e250c077a0b7c9bec481a1c41 (commit) via 090ba3670f4aeec3e95a81363c368076beae1d98 (commit) via a1316d3954d4873daf2b9e1fad504d97b45a4426 (commit) via 0d1d1c66c391745143e99894e0335d16c48123c4 (commit) via 68c47bb78209783fb875dd0278ede7ce1b84f91e (commit) from 05de2f5d858409b81e7b058362912316ac746e00 (commit) Those revisions listed above that are new to this repository have not appeared on any other notification email; so we list those revisions in full, below. - Log ----------------------------------------------------------------- commit c3223ffdfe4f8f5fe24d63aff21f17b61dbc57d5 Merge: 05de2f5d858409b81e7b058362912316ac746e00 1cf1f1609b11cb6a729b8d92ad882489a63d8b5c Author: Jonathan Knispel Date: Wed Nov 26 15:13:51 2008 +1100 Merge branch 'nathans' of git://oss.sgi.com:8090/nathans/pcp into dev Signed-off-by: Jonathan Knispel The commit message for 87bf1162aef138e7d9eec142f2cc446782511209 should have read "Remove README from LSRCFILES in memcache PMDA's GNUmakefile". commit 1cf1f1609b11cb6a729b8d92ad882489a63d8b5c Author: Nathan Scott Date: Wed Nov 26 13:47:18 2008 +1100 Ensure deb packages end up in build/deb not the top level. commit 3581a649887c7a38f3d4b92e666e38bc814d4650 Author: Nathan Scott Date: Wed Nov 26 12:41:16 2008 +1100 Fix "make; make clean; make" which failed in the include directory because the LDIRT there contained half of the configure-generated files. This makes us deal the same way with all configure-generated files, and: - make clobber/distclean => completely clean - make clean" => clean such that a rebuild will not need to run the configure script again ... which was the intention with these different targets. commit 87bf1162aef138e7d9eec142f2cc446782511209 Author: Jonathan Knispel Date: Wed Nov 26 08:46:49 2008 +1100 Remove README from memcache PMDA, this info is in the embedded Perl POD within the pmdamemcache.pl script now. commit ecc2a160303d414ea4a4e454e6606cdf42ea5863 Author: Nathan Scott Date: Tue Nov 25 18:12:25 2008 +1100 Add the hinv.pagesize and hinv.nfilesys metrics into pmdawindows. commit 0469b3aa229d83c2d08c608807bf0deb0810e77d Author: Ken McDonell Date: Tue Nov 25 16:53:16 2008 +1100 Add a missing endian conversion in pmdumplog "raw" mode, as reported here... On Tue, 2008-11-25 at 09:38 +1100, Nathan Scott wrote: > ... . Trying to get more info with -v was unhelpful - a > bug lurking here too perhaps? All archives produce this: > > $ pmdumplog -v 20081006.0 > Raw dump of physical archive file "20081006.0" ... > Dump ... record len: -2080374784 @ offset: 0 > Trailer botch: 36046160 != -2080374784 commit a552bc84f0ca5ff1a63b86a0183daa884ca4c802 Author: Nathan Scott Date: Tue Nov 25 16:43:54 2008 +1100 Switch to using PdhRemoveCounter as recommended by the Windows API documentation. commit 2bd9f0b05441fe0466da6bd9f0b384b24bb07fcf Author: Nathan Scott Date: Tue Nov 25 16:40:22 2008 +1100 Windows PMDA fixes. Fix some of the pmdaCache uses that were not passing in the indom identifier with the domain number OR'd in. Fix the M_REDO flag, so that instances which require full refresh are in fact refreshed on instance lookups. Finally, the network interface instance names now use the disk indom trick of replacing spaces with underscores to ensure the PCP instance domain rules are followed. commit bfab9527dddd7c3fbefa761706f47b70ff94bbb7 Author: Nathan Scott Date: Tue Nov 25 16:38:53 2008 +1100 Fix typo in Windows PMDA README file. commit e0ed285f0b6d23878a349b493482aa5fbf682434 Author: Nathan Scott Date: Tue Nov 25 10:24:47 2008 +1100 Update the makefiles for each of the Perl PMDAs to be able to generate man pages for each, and install them appropriately during the build. Also added man page POD docs to the MySQL PMDA, which was the last one missing a man page. commit 6ee1ccbf34b3e1e83c00374d4801e7c466efc780 Author: Nathan Scott Date: Tue Nov 25 10:21:16 2008 +1100 Add configure magic to find the pod2man utility, and some makefile macros to use it. This gives us the ability to generate installable man pages from inline "pod" docs in the Perl PMDAs. commit 7552b66acae654c9e1fe067622d7870dc6b536e7 Author: Nathan Scott Date: Mon Nov 24 21:51:41 2008 +1100 Update the standard domain numbers to avoid conflicts once more. Shift the Lustre PMDA to slot 93, to match the SGI tree and avoid the memcache PMDA. Also shunt the KVM PMDA down to slot 95, since it'd inadvertantly been added in at 93 as well. What a zoo! :) commit 1a10e27ed8d33c60012a2b0dab22d3c680de0385 Author: Nathan Scott Date: Mon Nov 24 10:03:17 2008 +1100 Manual pages for many of the Perl PMDAs. We should be able to invoke pod2man to generate those for installation, but thats a job for another day (as are the last couple of remaining Perl PMDAs needing better docs). commit 80245342a195ae2cad607c84e564cef76fe0fbc4 Author: Nathan Scott Date: Fri Nov 21 21:03:12 2008 +1100 Fix build when Infiniband PMDA is not configured, after last change. commit 32f96056e7f4e67e94da21f4b5a61f7d05bb749d Author: Nathan Scott Date: Fri Nov 21 08:11:24 2008 +1100 Do not unconditionally add the Infiniband PMDA into the set of DSOs which can be used via local context - only when configured into the build. Noticed while debugging under Win32. commit beb84d862142fc5e250c077a0b7c9bec481a1c41 Author: Nathan Scott Date: Fri Nov 21 08:08:48 2008 +1100 Cleanup PCP_VAR_DIR setting on Win32 so we dont see dup //s on path lookups. commit 090ba3670f4aeec3e95a81363c368076beae1d98 Author: Nathan Scott Date: Mon Nov 17 21:28:49 2008 +1100 Bump development snapshot version number. commit a1316d3954d4873daf2b9e1fad504d97b45a4426 Author: Nathan Scott Date: Mon Nov 17 20:06:39 2008 +1100 Pick a unique, unused domain number (93) for the KVM PMDA, and add kvm subdir into the build, only installed on Linux. commit 0d1d1c66c391745143e99894e0335d16c48123c4 Author: Nathan Scott Date: Mon Nov 17 08:57:33 2008 +1100 Initial version of the KVM PMDA, exporting metric values exposed by the Linux kernel virtualisation subsystem, KVM. commit 68c47bb78209783fb875dd0278ede7ce1b84f91e Merge: 7b2760b289b0a74a543c0be3dfdc17e4ab19483c 71625d6aab5ad7d45417d387c351b7a53470f76c Author: Nathan Scott Date: Fri Nov 14 16:08:22 2008 +1100 Merge branch 'dev' of git://oss.sgi.com/pcp/pcp into nathans ----------------------------------------------------------------------- Summary of changes: GNUmakefile | 3 +- Makepkgs | 7 +- VERSION.pcp | 2 +- configure.in | 9 +- debian/changelog | 9 +- src/include/GNUmakefile | 2 +- src/include/builddefs.in | 13 +- src/include/buildrules | 5 +- src/libpcp/src/GNUmakefile | 8 +- src/libpcp/src/connect.c | 2 + src/pmdas/GNUmakefile | 2 +- src/pmdas/dbping/GNUmakefile | 18 ++- src/pmdas/dbping/README | 51 ----- src/pmdas/dbping/dbprobe.pl | 19 ++- src/pmdas/dbping/pmdadbping.pl | 83 ++++++++- src/pmdas/{systemtap => kvm}/GNUmakefile | 24 ++- src/pmdas/{dbping => kvm}/Install | 15 +- src/pmdas/{mysql => kvm}/Remove | 4 +- src/pmdas/kvm/pmdakvm.pl | 110 +++++++++++ src/pmdas/memcache/GNUmakefile | 21 ++- src/pmdas/memcache/README | 52 ----- src/pmdas/memcache/pmdamemcache.pl | 137 ++++++++++---- src/pmdas/mysql/GNUmakefile | 17 ++- src/pmdas/mysql/pmdamysql.pl | 81 ++++++++ src/pmdas/systemtap/GNUmakefile | 16 ++- src/pmdas/systemtap/pmdasystemtap.pl | 82 ++++++++ src/pmdas/vmware/GNUmakefile | 14 ++- src/pmdas/vmware/pmdavmware.pl | 77 ++++++++ src/pmdas/windows/README | 2 - src/pmdas/windows/hypnotoad.h | 5 +- src/pmdas/windows/instance.c | 30 ++-- src/pmdas/windows/libpdh.def | 1 + src/pmdas/windows/libpdh.h | 3 + src/pmdas/windows/open.c | 302 +++++++++++++++--------------- src/pmdas/windows/pmda.c | 31 +++- src/pmdas/windows/pmns.hinv | 2 + src/pmdumplog/pmdumplog.c | 6 +- src/pmns/stdpmid.add-on | 2 +- src/pmns/stdpmid.pcp | 1 + 39 files changed, 890 insertions(+), 378 deletions(-) delete mode 100644 src/pmdas/dbping/README copy src/pmdas/{systemtap => kvm}/GNUmakefile (74%) copy src/pmdas/{dbping => kvm}/Install (78%) copy src/pmdas/{mysql => kvm}/Remove (96%) create mode 100644 src/pmdas/kvm/pmdakvm.pl delete mode 100644 src/pmdas/memcache/README -- Performance Co-Pilot (PCP) From kenj@internode.on.net Wed Nov 26 03:18:11 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=-2.5 required=5.0 tests=J_CHICKENPOX_44, J_CHICKENPOX_45,J_CHICKENPOX_56,J_CHICKENPOX_64,J_CHICKENPOX_65, LOCAL_GNU_PATCH autolearn=unavailable version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAQ9IARf031278 for ; Wed, 26 Nov 2008 03:18:11 -0600 X-ASG-Debug-ID: 1227691086-48b201870000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from mail.internode.on.net (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 1A2801604029 for ; Wed, 26 Nov 2008 01:18:07 -0800 (PST) Received: from mail.internode.on.net (bld-mail06.adl2.internode.on.net [203.16.214.70]) by cuda.sgi.com with ESMTP id mjtz9fU3kF4jA9HB for ; Wed, 26 Nov 2008 01:18:07 -0800 (PST) Received: from [192.168.1.100] (unverified [118.208.222.17]) by mail.internode.on.net (SurgeMail 3.8f2) with ESMTP id 35121968-1927428 for ; Wed, 26 Nov 2008 19:48:02 +1030 (CDT) X-ASG-Orig-Subj: pcpqa patch Subject: pcpqa patch From: Ken McDonell Reply-To: kenj@internode.on.net To: pcp@oss.sgi.com Content-Type: multipart/mixed; boundary="=-eLWO3st/5zBNp7GM3PDX" Date: Wed, 26 Nov 2008 20:17:43 +1100 Message-Id: <1227691063.2540.30.camel@bozo> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 X-Barracuda-Connect: bld-mail06.adl2.internode.on.net[203.16.214.70] X-Barracuda-Start-Time: 1227691089 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -1.52 X-Barracuda-Spam-Status: No, SCORE=-1.52 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests=BSF_RULE7568M X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11191 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- 0.50 BSF_RULE7568M Custom Rule 7568M --=-eLWO3st/5zBNp7GM3PDX Content-Type: text/plain Content-Transfer-Encoding: 7bit This patch goes the the associated pmlogger_daily patch and contains a few other fixes. I don't seem to be able to get the new files into the patch so they are attached separately. And these files need to be removed/added/renamed from the repository ... # deleted: 151.out.00-01 # deleted: 151.out.01-24 # new file: 338 # new file: 338.out # renamed: 530.out.02-24 -> 530.out # deleted: 530.out.00-01 # deleted: 530.out.01-02 # deleted: 532.out.00-01 # deleted: 532.out.01-24 Here's some terse take messages for the patched files ... 151 338 [new] 530 530.out 532 564 Reworked entirely for pmlogger_daily changes to handle daylight savings changes. 183 183.out.linux Improved filtering of kill messages. 202 Better "full" diagnostics to help debug tricky gcc problem! 282 282.out.3 Some residual pmlogger/control parametrisation changes. Updated after pmlogger_daily changes to handle daylight savings changes. Handle some slightly different messages from the pcp init scripts in error conditions. check Minor tweak for output when test from "groups" is in fact missing (like all the sgi ones in open source land). common.filter Residual pmlogger/control parametrisation change. group owner Add 338. show-me Fix syntax botch. --=-eLWO3st/5zBNp7GM3PDX Content-Disposition: attachment; filename=pcpqa.patch Content-Type: text/x-patch; name=pcpqa.patch; charset=UTF-8 Content-Transfer-Encoding: 7bit diff --git a/151 b/151 index 20891e4..ec60ae7 100755 --- a/151 +++ b/151 @@ -19,7 +19,7 @@ echo "QA output created by $seq" . ./common.filter . ./common.check -# check for new version ... may need -o to force "old style" behaviour +# check for new version ... # $PCP_BINADM_DIR/pmlogger_daily -\? 2>&1 >/tmp/eek if $PCP_BINADM_DIR/pmlogger_daily -\? 2>&1 | grep '^.*-o.*(old style)' >/dev/null @@ -31,21 +31,6 @@ else exit 0 fi -# Because pmlogger_daily is way too clever to use the same criteria -# for deciding which logs to merge and which to cull, we have to have -# QA output in 2 different times of the day - -case "`date +%H`" in -00) - # just after midnight - rm -f $seq.out && ln $seq.out.00-01 $seq.out || exit 1 - ;; -*) - # any other time - rm -f $seq.out && ln $seq.out.01-24 $seq.out || exit 1 - ;; -esac - _filter() { tee -a $seq.full \ @@ -77,27 +62,39 @@ mkdir $tmp mkdir $tmp/foo # real QA test starts here -verbose=-V + echo >$tmp/sed -for day in 0 1 2 3 4 5 +# unmerged archives for these days ... +# +for day in 0 1 2 3 do - fmt=%Y%m%d - for hour in 1 4 16 do [ $day -gt 0 -a $hour -gt 1 ] && continue - - namea=`pmdate -${day}d -${hour}H $fmt.%H.%M` - nameb=`pmdate -${day}d -${hour}H -1M $fmt.%H.%M` - namec=`pmdate -${day}d -${hour}H -2M $fmt.%H.%M` + # aiming for the time to be hour:10, hour:20 and hour:30 on + # each day, even if that is ahead of NOW for today + # + now_hr=`pmdate %H` + hr_set=`expr $hour - $now_hr` + [ "$hr_set" -ge 0 ] && hr_set="+$hr_set" + now_min=`pmdate %M` + min_set10=`expr 10 - $now_min` + [ "$min_set10" -ge 0 ] && min_set10="+$min_set10" + min_set20=`expr 20 - $now_min` + [ "$min_set20" -ge 0 ] && min_set20="+$min_set20" + min_set30=`expr 30 - $now_min` + [ "$min_set30" -ge 0 ] && min_set30="+$min_set30" + namea=`pmdate -${day}d ${hr_set}H ${min_set10}M %Y%m%d.%H.%M` + nameb=`pmdate -${day}d ${hr_set}H ${min_set20}M %Y%m%d.%H.%M` + namec=`pmdate -${day}d ${hr_set}H ${min_set30}M %Y%m%d.%H.%M` stampa=`echo $namea | sed -e 's/\.//g'` stampb=`echo $nameb | sed -e 's/\.//g'` stampc=`echo $namec | sed -e 's/\.//g'` nameb=${namea}-00 namec=${namea}-01 - echo "s/$namea/TODAY-$day.-${hour}H.MM/g" >>$tmp/sed - echo "s/$nameb/TODAY-$day.-${hour}H.MM-00/g" >>$tmp/sed - echo "s/$namec/TODAY-$day.-${hour}H.MM-01/g" >>$tmp/sed + echo "s/$namea/TODAY-$day.${hour}H.MM/g" >>$tmp/sed + echo "s/$nameb/TODAY-$day.${hour}H.MM-00/g" >>$tmp/sed + echo "s/$namec/TODAY-$day.${hour}H.MM-01/g" >>$tmp/sed for ext in 0 index meta do @@ -112,17 +109,36 @@ do touch -t $stampc $tmp/$namec.$ext done done + echo "s/`pmdate -${day}d ${hr_set}H ${min_set10}M %Y%m%d`/TODAY-$day/g" >>$tmp/sed done -( cd $tmp; tar cf - *.0 *.meta *.index ) | ( cd $tmp/foo; tar xpf - ) +# merged archives for these days ... +# +for day in 4 5 6 +do + # aiming for the time to be 00:10 like when cron does its stuff + # + hr_set=-`pmdate %H` + now_min=`pmdate %M` + min_set=`expr 10 - $now_min` + [ "$min_set" -ge 0 ] && min_set="+$min_set" + namea=`pmdate -${day}d ${hr_set}H ${min_set}M %Y%m%d` + stampa=`echo $namea | sed -e 's/\.//g'`0010 + echo "s/$namea/TODAY-$day/g" >>$tmp/sed + + for ext in 0 index meta + do + cp src-oss/arch_a.$ext $tmp/$namea.$ext + touch -t $stampa $tmp/$namea.$ext + done + echo "s/$namea/TODAY-$day/g" >>$tmp/sed +done echo "s/TODAY-0/TODAY/g" >>$tmp/sed echo "s/`pmdate %Y%m%d`/TODAY/g" >>$tmp/sed -echo "s/`pmdate -1d %Y%m%d`/TODAY-1/g" >>$tmp/sed -echo "s/`pmdate -2d %Y%m%d`/TODAY-2/g" >>$tmp/sed -echo "s/`pmdate -3d %Y%m%d`/TODAY-3/g" >>$tmp/sed -echo "s/`pmdate -4d %Y%m%d`/TODAY-4/g" >>$tmp/sed -echo "s/`pmdate -5d %Y%m%d`/TODAY-5/g" >>$tmp/sed + +( cd $tmp; tar cf - *.0 *.meta *.index 2>/dev/null ) \ +| ( cd $tmp/foo; tar xpf - 2>/dev/null ) echo >$tmp/control '$version=1.1 LOCALHOSTNAME n n '"$tmp"' ./config' @@ -130,13 +146,18 @@ LOCALHOSTNAME n n '"$tmp"' ./config' echo echo "Before ..." ls $tmp | sed -f $tmp/sed | LC_COLLATE=POSIX sort | tee $tmp.before +echo "sed ..." >>$seq.full +cat $tmp/sed >>$seq.full +echo "before ..." >>$seq.full ls -l $tmp >>$seq.full # list of culled files is in directory order, hence the cut and sort # complexity at the end of this pipeline # echo -pmlogger_daily -c $tmp/control -k 4 $verbose 2>&1 \ +echo "pmlogger_daily -k 4 -V ..." >>$seq.full +pmlogger_daily -c $tmp/control -k 4 -V 2>&1 \ +| tee -a $tmp.full \ | _filter \ | $PCP_AWK_PROG ' state == 0 { print >"'$tmp.out.0'" @@ -145,17 +166,19 @@ state == 0 { print >"'$tmp.out.0'" } state == 1 && NF == 0 { state = 2 } state == 1 { for (i = 1; i <= NF; i++) print $i >"'$tmp.out.1'" }' -cat $tmp.out.0 -LC_COLLATE=POSIX sort $tmp.out.1 +[ -f $tmp.out.0 ] && cat $tmp.out.0 +[ -f $tmp.out.1 ] && LC_COLLATE=POSIX sort $tmp.out.1 echo "out.0 ..." >>$seq.full -cat $tmp.out.0 >>$seq.full +[ -f $tmp.out.0 ] && cat $tmp.out.0 >>$seq.full echo "out.1 ..." >>$seq.full -cat $tmp.out.1 >>$seq.full +[ -f $tmp.out.1 ] && cat $tmp.out.1 >>$seq.full echo echo "After ..." ls $tmp | sed -f $tmp/sed | LC_COLLATE=POSIX sort | tee $tmp.after +echo "after ..." >>$seq.full +ls -l $tmp >>$seq.full echo echo "Differences ..." diff --git a/183 b/183 index 8bc85a9..8f5e455 100755 --- a/183 +++ b/183 @@ -195,19 +195,18 @@ fi sleep 2 pmnewlog -p $pid $tmp 2>&1 \ +| tee -a $seq.full \ | _filter \ | sed \ -e "/\.save/s/.*/-- ps line removed --/" \ - -e "s/$pid/PID/" \ - -e '/kill/s/: line [0-9][0-9]*:/:/' \ -e '/^kill:.*Not owner/s/Not owner/Operation not permitted/' \ -e '/.*: kill:.*Not owner/s/Not owner/Operation not permitted/' \ + -e 's/.*kill:.*Operation not permitted/KILL: Operation not permitted/' \ -e '/to get logging state/d' \ -e '/errors from talking/d' \ -e '/continuing/d' \ -e '/Reason: Metric not supported by this version/d' \ - -e '/^$/d' \ - -e 's/kill: PID: p/P/' + -e '/^$/d' ./sudo kill -INT $pid diff --git a/183.out.linux b/183.out.linux index 93c96d4..6019411 100644 --- a/183.out.linux +++ b/183.out.linux @@ -70,8 +70,8 @@ Most likely this pmlogger instance is inactive. Sorry, but this is fatal. No new pmlogger instance has been started. === wrong uid to kill pmlogger === -PCP_BINADM_DIR/pmnewlog: kill: (PID) - Operation not permitted -PCP_BINADM_DIR/pmnewlog: kill: (PID) - Operation not permitted -PCP_BINADM_DIR/pmnewlog: kill: (PID) - Operation not permitted +KILL: Operation not permitted +KILL: Operation not permitted +KILL: Operation not permitted failed! Sorry, but this is fatal. No new pmlogger instance has been started. diff --git a/202 b/202 index 878ef6a..4f6465a 100755 --- a/202 +++ b/202 @@ -15,10 +15,13 @@ echo "QA output created by $seq" tmp=/tmp/$$ trap "rm -f $tmp.*; exit" 0 1 2 3 15 +#debug# tmp=`pwd`/tmp +rm -f $tmp.* $seq.full _filter() { pmdumplog $1 \ + | tee -a $seq.full \ | $PCP_AWK_PROG ' /\/ { mark++; next } /^[0-9][0-9]:[0-9][0-9]:[0-9][0-9]\./ { rec++ } @@ -34,17 +37,23 @@ END { printf "%d records",rec # real QA test starts here rm -f $tmp.A.* -echo "log mandatory on 20 msec sample.milliseconds" \ -| pmlogger -s 4 $tmp.A >/dev/null 2>&1 +echo "log mandatory on 40 msec sample.milliseconds" \ +| pmlogger -s 4 -l $tmp.log $tmp.A >>$seq.full 2>&1 echo echo log.A +cat $tmp.log >>$seq.full +ls -l $tmp.A.* >>$seq.full +pmdumplog -l $tmp.A >>$seq.full _filter $tmp.A rm -f $tmp.B.* -echo "log mandatory on 25 msec sample.milliseconds" \ -| pmlogger -s 6 $tmp.B >/dev/null 2>&1 +echo "log mandatory on 50 msec sample.milliseconds" \ +| pmlogger -s 6 -l $tmp.log $tmp.B >>$seq.full 2>&1 echo echo log.B +cat $tmp.log >>$seq.full +ls -l $tmp.B.* >>$seq.full +pmdumplog -l $tmp.A >>$seq.full _filter $tmp.B rm -f $tmp.C.* @@ -62,7 +71,7 @@ cat $tmp.A.stamp $tmp.B.stamp \ echo echo "pmval [no interpolate] ..." pmval -U $tmp.C sample.milliseconds 2>&1 \ -| tee $seq.full \ +| tee -a $seq.full \ | $PCP_AWK_PROG ' /suspended/ { mark++; next } /^[0-9][0-9]:[0-9][0-9]:[0-9][0-9]\./ { rec++ } diff --git a/282 b/282 index 820d049..1449856 100755 --- a/282 +++ b/282 @@ -94,6 +94,8 @@ _poke_pmcd() -e 's/\(install the pcp.sw.base\) .*/\1 ... PCP 2.0 or later .../' \ -e 's/.*\(consult PMLOGGER.CHECK(1), else\)/ ... \1/' \ -e '/root.*.PCP_BINADM_DIR.pmcd/s/.*/-- ps output line --/' \ + -e '/^failed (status=1)/d' \ + -e '/^Waiting for PMCD/s/\.\.\.[. ]*$/DOTS/' \ -e '/^$/d' can_wait=5 @@ -167,7 +169,7 @@ _save_setup() { for file in $PCP_PMCDCONF_PATH $SYSPMDA \ $PCP_PMCD_PROG \ - $CHECK $PCP_VAR_DIR/config/pmlogger/control \ + $CHECK $PCP_PMLOGGERCONTROL_PATH \ $PCP_PMLOGGEROPTIONS_PATH do [ -f $file.$$ ] && $sudo rm -f $file.$$ @@ -184,7 +186,7 @@ _cleanup() echo "Restore the state of the universe ..." for file in $PCP_PMCDCONF_PATH $SYSPMDA \ $PCP_PMCD_PROG \ - $CHECK $PCP_VAR_DIR/config/pmlogger/control \ + $CHECK $PCP_PMLOGGERCONTROL_PATH \ $PCP_PMLOGGEROPTIONS_PATH do [ -f $file.$$ ] && $sudo mv $file.$$ $file @@ -289,7 +291,7 @@ End-of-File echo echo "=== expect error: pmlogger/control moved, so no pmlogger ===" - $sudo rm -f $PCP_VAR_DIR/config/pmlogger/control + $sudo rm -f $PCP_PMLOGGERCONTROL_PATH _poke_pmcd_n_logger if [ $PCP_PLATFORM = irix ] @@ -299,20 +301,20 @@ End-of-File # echo echo "=== expect warning: host not in pmlogger/control, so initialize ===" - $sudo rm -f $PCP_VAR_DIR/config/pmlogger/control - sed -e "/LOCALHOSTNAME/d" -e "/`hostname`/d" <$PCP_VAR_DIR/config/pmlogger/control.$$ >$tmp.out - $sudo cp $tmp.out $PCP_VAR_DIR/config/pmlogger/control + $sudo rm -f $PCP_PMLOGGERCONTROL_PATH + sed -e "/LOCALHOSTNAME/d" -e "/`hostname`/d" <$PCP_PMLOGGERCONTROL_PATH.$$ >$tmp.out + $sudo cp $tmp.out $PCP_PMLOGGERCONTROL_PATH sleepy=true _poke_pmcd_n_logger sleepy=false # added by mg - $sudo cp $PCP_VAR_DIR/config/pmlogger/control.$$ $PCP_VAR_DIR/config/pmlogger/control + $sudo cp $PCP_PMLOGGERCONTROL_PATH.$$ $PCP_PMLOGGERCONTROL_PATH echo echo "=== expect pmlogger connect failure: same case, but pmlogger.options is bogus ===" - $sudo rm -f $PCP_VAR_DIR/config/pmlogger/control - sed -e "/LOCALHOSTNAME/d" -e "/`hostname`/d" <$PCP_VAR_DIR/config/pmlogger/control.$$ >$tmp.out - $sudo cp $tmp.out $PCP_VAR_DIR/config/pmlogger/control + $sudo rm -f $PCP_PMLOGGERCONTROL_PATH + sed -e "/LOCALHOSTNAME/d" -e "/`hostname`/d" <$PCP_PMLOGGERCONTROL_PATH.$$ >$tmp.out + $sudo cp $tmp.out $PCP_PMLOGGERCONTROL_PATH echo "-c /no/such/config" >$tmp.tmp $sudo mv $tmp.tmp $PCP_PMLOGGEROPTIONS_PATH _poke_pmcd_n_logger \ @@ -324,7 +326,7 @@ End-of-File -e '/^total /s/.*/.../' \ -e '/missing$/s/[0-9][0-9][0-9][0-9][0-9][0-9][0-9][0-9]\.[0-9][0-9]\.[0-9][0-9]/CHECK/' $sudo rm -f $PCP_PMLOGGEROPTIONS_PATH - $sudo cp $PCP_VAR_DIR/config/pmlogger/control.$$ $PCP_VAR_DIR/config/pmlogger/control + $sudo cp $PCP_PMLOGGERCONTROL_PATH.$$ $PCP_PMLOGGERCONTROL_PATH fi # no chkconfig verbose for Linux diff --git a/282.out.3 b/282.out.3 index 1f10d04..f7131b3 100644 --- a/282.out.3 +++ b/282.out.3 @@ -1,12 +1,12 @@ QA output created by 282 === expect error: no $PCP_PMCDCONF_PATH === -Waiting for PMCD to terminate ... +Waiting for PMCD to terminate DOTS $PCP_RC_DIR/pcp: -Warning: PMCD control file $PCP_VAR_DIR/config/pmcd/pmcd.conf is missing, cannot identify PMDAs +Warning: PMCD control file $PCP_PMCDCONF_PATH is missing, cannot identify PMDAs to be terminated. $PCP_RC_DIR/pcp: -Error: PMCD control file $PCP_VAR_DIR/config/pmcd/pmcd.conf is missing, cannot start PMCD. +Error: PMCD control file $PCP_PMCDCONF_PATH is missing, cannot start PMCD. === probe pmcd: pmprobe: Cannot connect to PMCD on host "HOST": Connection refused === simple case, no frills, no errors === @@ -20,16 +20,17 @@ $PCP_RC_DIR/pcp: Error: You must be root (uid 0) to start or stop the Performance Co-Pilot PMCD. === probe pmcd: OK === start again === -Waiting for PMCD to terminate ... +Waiting for PMCD to terminate DOTS Performance Co-Pilot starting PMCD (logfile is $PCP_LOG_DIR/pmcd.log) ... Performance Co-Pilot starting archive loggers ... === probe pmcd: OK === probe pmlogger: OK === expect 2 x connect failure: pmcd is true(1), so hard for pmlogger === -Waiting for PMCD to terminate ... +Waiting for PMCD to terminate DOTS Performance Co-Pilot starting PMCD (logfile is $PCP_LOG_DIR/pmcd.log) ... Performance Co-Pilot starting archive loggers ... +$PCP_RC_DIR/pcp: pmcd_wait failed: exit status: 2 === probe pmcd: pmprobe: Cannot connect to PMCD on host "HOST": Connection refused === probe pmlogger: Unable to connect to primary pmlogger at HOST: Connection refused === start again === @@ -38,7 +39,7 @@ Performance Co-Pilot starting archive loggers ... === probe pmcd: OK === probe pmlogger: OK === expect difficulty terminating pmcd ... it has been sent a SIGSTOP === -Waiting for PMCD to terminate ........... +Waiting for PMCD to terminate DOTS Process ... -- ps output line -- $PCP_RC_DIR/pcp: Warning: Forcing PMCD to terminate! @@ -47,7 +48,7 @@ Performance Co-Pilot starting archive loggers ... === probe pmcd: OK === expect warning: PMLOGGER.CHECK moved, so no pmlogger === -Waiting for PMCD to terminate ... +Waiting for PMCD to terminate DOTS Performance Co-Pilot starting PMCD (logfile is $PCP_LOG_DIR/pmcd.log) ... $PCP_RC_DIR/pcp: Warning: Performance Co-Pilot installation is incomplete (at least the @@ -57,11 +58,11 @@ Warning: Performance Co-Pilot installation is incomplete (at least the === probe pmlogger: Unable to connect to primary pmlogger at HOST: Connection refused === expect error: pmlogger/control moved, so no pmlogger === -Waiting for PMCD to terminate ... +Waiting for PMCD to terminate DOTS Performance Co-Pilot starting PMCD (logfile is $PCP_LOG_DIR/pmcd.log) ... Performance Co-Pilot starting archive loggers ... $PCP_RC_DIR/pcp: -Error: PCP archive logger control file $PCP_VAR_DIR/config/pmlogger/control +Error: PCP archive logger control file $PCP_PMLOGGERCONTROL_PATH is missing! Cannot start any Performance Co-Pilot archive logger(s). === probe pmcd: OK === probe pmlogger: Unable to connect to primary pmlogger at HOST: Connection refused diff --git a/338 b/338 index 7a2d063..04a325f 100755 --- a/338 +++ b/338 @@ -56,7 +56,6 @@ END { if ( eatlines ) { -e "s;$tmp;TMP;g" \ -e "s/`hostname`/HOST/g" \ -e "s/`hostname | sed -e 's/\..*//'`/HOST/g" \ - -e '/Possibly a daylight saving change/d' \ -e '/\/var\/adm\/pcplog\/NOTICES/d' \ | $PCP_AWK_PROG ' /-rw-r--r--/ { print $1,"...",$5,"...",$9; next } @@ -94,9 +93,12 @@ do # now_hr=`pmdate %H` hra=`expr 3 - $now_hr` + [ "$hra" -ge 0 ] && hra="+$hra" hrb=`expr 5 - $now_hr` + [ "$hrb" -ge 0 ] && hrb="+$hrb" now_min=`pmdate %M` min=`expr 10 - $now_min` + [ "$min" -ge 0 ] && min="+$min" namea=`pmdate -${day}d ${hra}H ${min}M $fmt.%H.%M` nameb=`pmdate -${day}d ${hrb}H ${min}M $fmt.%H.%M` yesterday=`pmdate -1d ${hrb}H ${min}M $fmt` diff --git a/530 b/530 index 29b651d..73487b6 100755 --- a/530 +++ b/530 @@ -2,8 +2,8 @@ # PCP QA Test No. 530 # $Revision: 1.13 $ # pmlogger_daily ... do you work at all? -# dual of 314 ... this is the PCP 2.1 version # +# OLD VERSION for YYDDMM style naming convention # # Copyright (c) 1995-2002 Silicon Graphics, Inc. All Rights Reserved. # @@ -11,25 +11,6 @@ owner=kenmcd seq=`basename $0` -# Because pmlogger_daily is way too clever to use the same criteria -# for deciding which logs to merge and which to cull, we have to have -# QA output in 3 different times of the day - -case "`date +%H`" in -00) - # just after midnight - rm -f $seq.out && ln $seq.out.00-01 $seq.out || exit 1 - ;; -01) - # just after 1am - rm -f $seq.out && ln $seq.out.01-02 $seq.out || exit 1 - ;; -*) - # any other time of the day - rm -f $seq.out && ln $seq.out.02-24 $seq.out || exit 1 - ;; -esac - echo "QA output created by $seq" # check for new version ... may need -o to force "old style" behaviour @@ -74,7 +55,6 @@ END { if ( eatlines ) { -e "s;$tmp;TMP;g" \ -e "s/`hostname`/HOST/g" \ -e "s/`hostname | sed -e 's/\..*//'`/HOST/g" \ - -e '/Possibly a daylight saving change/d' \ -e '/\/var\/adm\/pcplog\/NOTICES/d' \ | $PCP_AWK_PROG ' /-rw-r--r--/ { print $1,"...",$5,"...",$9; next } @@ -103,36 +83,39 @@ do echo >$tmp/sed for day in 0 1 2 3 4 5 do - for fmt in %Y%m%d %y%m%d + fmt=%y%m%d + # aiming for the time to be 03:10 and 05:10 on each day, + # even if that is ahead of NOW for today + # + now_hr=`pmdate %H` + hra=`expr 3 - $now_hr` + [ "$hra" -ge 0 ] && hra="+$hra" + hrb=`expr 5 - $now_hr` + [ "$hrb" -ge 0 ] && hrb="+$hrb" + now_min=`pmdate %M` + min=`expr 10 - $now_min` + [ "$min" -ge 0 ] && min="+$min" + namea=`pmdate -${day}d ${hra}H ${min}M $fmt.%H.%M` + nameb=`pmdate -${day}d ${hrb}H ${min}M $fmt.%H.%M` + yesterday=`pmdate -1d ${hrb}H ${min}M $fmt` + stampa=`echo $namea | sed -e 's/\.//g'` + stampb=`echo $nameb | sed -e 's/\.//g'` + echo "s/$namea/NOW-$day.03.10/g" >>$tmp/sed + echo "s/$nameb/NOW-$day.05.10/g" >>$tmp/sed + + for ext in 0 index meta do - namea=`pmdate -${day}d -1H $fmt.%H.%M` - nameb=`pmdate -${day}d -2H $fmt.%H.%M` - stampa=`echo $namea | sed -e 's/\.//g'` - stampb=`echo $nameb | sed -e 's/\.//g'` - if [ "$fmt" = "%Y%m%d" ] - then - echo "s/$namea/NOW-$day.-1H.MM/g" >>$tmp/sed - echo "s/$nameb/NOW-$day.-2H.MM/g" >>$tmp/sed - else - echo "s/$namea/NOW-$day-old.-1H.MM/g" >>$tmp/sed - echo "s/$nameb/NOW-$day-old.-2H.MM/g" >>$tmp/sed - fi - - for ext in 0 index meta - do - cp src-oss/arch_a.$ext $tmp/$namea.$ext - touch -t $stampa $tmp/$namea.$ext - cp src-oss/arch_b.$ext $tmp/$nameb.$ext - touch -t $stampb $tmp/$nameb.$ext - done + cp src-oss/arch_a.$ext $tmp/$namea.$ext + touch -t $stampa $tmp/$namea.$ext + cp src-oss/arch_b.$ext $tmp/$nameb.$ext + touch -t $stampb $tmp/$nameb.$ext done done ls -lt $tmp >>$seq.full echo "s/NOW-0/TODAY/g" >>$tmp/sed echo "s/NOW-1/YESTERDAY/g" >>$tmp/sed - echo "s/`pmdate %Y%m%d`/TODAY/g" >>$tmp/sed - echo "s/`pmdate -1d %Y%m%d`/YESTERDAY/g" >>$tmp/sed - echo "s/`pmdate -1d %y%m%d`/YESTERDAY-old/g" >>$tmp/sed + echo "s/`pmdate %y%m%d`/TODAY/g" >>$tmp/sed + echo "s/$yesterday/YESTERDAY/g" >>$tmp/sed echo >$tmp/control '$version=1.1 LOCALHOSTNAME n n '"$tmp"' ./config' @@ -140,15 +123,18 @@ LOCALHOSTNAME n n '"$tmp"' ./config' echo echo "=== [${verbose+verbose}] regular show ===" echo "=== [${verbose+verbose}] regular show ===" >>$seq.full - pmlogger_daily $oflag -c $tmp/control -k 2 $verbose 2>&1 \ + pmlogger_daily $oflag -c $tmp/control -k 3 $verbose 2>&1 \ | _filter ls $tmp | sed -f $tmp/sed | LC_COLLATE=POSIX sort echo echo "=== [${verbose+verbose}] nothing to do but cull some more ===" echo "=== [${verbose+verbose}] nothing to do but cull some more ===" >>$seq.full - rm -f $tmp/`pmdate -${day}d %Y%m%d`.* - pmlogger_daily $oflag -c $tmp/control -k 1 $verbose 2>&1 \ + for ext in 0 index meta + do + mv $tmp/$yesterday.$ext $tmp/save.$ext + done + pmlogger_daily $oflag -c $tmp/control -k 2 $verbose 2>&1 \ | _filter ls $tmp | sed -f $tmp/sed | LC_COLLATE=POSIX sort @@ -157,10 +143,10 @@ LOCALHOSTNAME n n '"$tmp"' ./config' echo "=== [${verbose+verbose}] error, output already exists ===" >>$seq.full for ext in 0 index meta do - cp src-oss/arch_a.$ext $tmp/`pmdate -1d %Y%m%d`.17.18.$ext - cp src-oss/arch_b.$ext $tmp/`pmdate -1d %y%m%d`.18.19.$ext + cp src-oss/arch_a.$ext $tmp/$yesterday.17.18.$ext + mv $tmp/save.$ext $tmp/$yesterday.$ext done - pmlogger_daily $oflag -c $tmp/control -k 1 $verbose 2>&1 \ + pmlogger_daily $oflag -c $tmp/control -k 2 $verbose 2>&1 \ | _filter ls $tmp | sed -f $tmp/sed | LC_COLLATE=POSIX sort diff --git a/530.out b/530.out index eadf0c4..4d0ba4d 100644 --- a/530.out +++ b/530.out @@ -3,30 +3,24 @@ QA output created by 530 === [] regular show === PMLOGGER.DAILY: Error: no pmlogger instance running for host "HOST" [TMP/control:2] ... logging for host "HOST" unchanged -NOW-2-old.-1H.MM.0 -NOW-2-old.-1H.MM.index -NOW-2-old.-1H.MM.meta -NOW-2-old.-2H.MM.0 -NOW-2-old.-2H.MM.index -NOW-2-old.-2H.MM.meta -NOW-2.-1H.MM.0 -NOW-2.-1H.MM.index -NOW-2.-1H.MM.meta -NOW-2.-2H.MM.0 -NOW-2.-2H.MM.index -NOW-2.-2H.MM.meta -YESTERDAY-old.-1H.MM.0 -YESTERDAY-old.-1H.MM.index -YESTERDAY-old.-1H.MM.meta -YESTERDAY-old.-2H.MM.0 -YESTERDAY-old.-2H.MM.index -YESTERDAY-old.-2H.MM.meta -YESTERDAY.-1H.MM.0 -YESTERDAY.-1H.MM.index -YESTERDAY.-1H.MM.meta -YESTERDAY.-2H.MM.0 -YESTERDAY.-2H.MM.index -YESTERDAY.-2H.MM.meta +NOW-2.03.10.0 +NOW-2.03.10.index +NOW-2.03.10.meta +NOW-2.05.10.0 +NOW-2.05.10.index +NOW-2.05.10.meta +NOW-3.03.10.0 +NOW-3.03.10.index +NOW-3.03.10.meta +NOW-3.05.10.0 +NOW-3.05.10.index +NOW-3.05.10.meta +TODAY.03.10.0 +TODAY.03.10.index +TODAY.03.10.meta +TODAY.05.10.0 +TODAY.05.10.index +TODAY.05.10.meta YESTERDAY.0 YESTERDAY.index YESTERDAY.meta @@ -36,22 +30,22 @@ sed === [] nothing to do but cull some more === PMLOGGER.DAILY: Error: no pmlogger instance running for host "HOST" [TMP/control:2] ... logging for host "HOST" unchanged -YESTERDAY-old.-1H.MM.0 -YESTERDAY-old.-1H.MM.index -YESTERDAY-old.-1H.MM.meta -YESTERDAY-old.-2H.MM.0 -YESTERDAY-old.-2H.MM.index -YESTERDAY-old.-2H.MM.meta -YESTERDAY.-1H.MM.0 -YESTERDAY.-1H.MM.index -YESTERDAY.-1H.MM.meta -YESTERDAY.-2H.MM.0 -YESTERDAY.-2H.MM.index -YESTERDAY.-2H.MM.meta -YESTERDAY.0 -YESTERDAY.index -YESTERDAY.meta +NOW-2.03.10.0 +NOW-2.03.10.index +NOW-2.03.10.meta +NOW-2.05.10.0 +NOW-2.05.10.index +NOW-2.05.10.meta +TODAY.03.10.0 +TODAY.03.10.index +TODAY.03.10.meta +TODAY.05.10.0 +TODAY.05.10.index +TODAY.05.10.meta control +save.0 +save.index +save.meta sed === [] error, output already exists === @@ -59,21 +53,18 @@ PMLOGGER.DAILY: Error: no pmlogger instance running for host "HOST" [TMP/control:2] ... logging for host "HOST" unchanged PMLOGGER.DAILY: Warning: output archive (YESTERDAY) already exists [TMP/control:2] ... skip log merging, culling and compressing for host "HOST" -YESTERDAY-old.-1H.MM.0 -YESTERDAY-old.-1H.MM.index -YESTERDAY-old.-1H.MM.meta -YESTERDAY-old.-2H.MM.0 -YESTERDAY-old.-2H.MM.index -YESTERDAY-old.-2H.MM.meta -YESTERDAY-old.18.19.0 -YESTERDAY-old.18.19.index -YESTERDAY-old.18.19.meta -YESTERDAY.-1H.MM.0 -YESTERDAY.-1H.MM.index -YESTERDAY.-1H.MM.meta -YESTERDAY.-2H.MM.0 -YESTERDAY.-2H.MM.index -YESTERDAY.-2H.MM.meta +NOW-2.03.10.0 +NOW-2.03.10.index +NOW-2.03.10.meta +NOW-2.05.10.0 +NOW-2.05.10.index +NOW-2.05.10.meta +TODAY.03.10.0 +TODAY.03.10.index +TODAY.03.10.meta +TODAY.05.10.0 +TODAY.05.10.index +TODAY.05.10.meta YESTERDAY.0 YESTERDAY.17.18.0 YESTERDAY.17.18.index @@ -91,51 +82,35 @@ PMLOGGER.DAILY: Error: no pmlogger instance running for host "HOST" [TMP/control:2] ... logging for host "HOST" unchanged Input archives to be merged: - TODAY-old.-2H.MM - TODAY-old.-1H.MM - TODAY.-2H.MM - TODAY.-1H.MM + YESTERDAY.03.10 + YESTERDAY.05.10 Output archive files: --rw-r--r-- ... 537 ... YESTERDAY.meta +-rw-r--r-- ... 455 ... YESTERDAY.meta -rw-r--r-- ... 192 ... YESTERDAY.index --rw-r--r-- ... 1596 ... YESTERDAY.0 +-rw-r--r-- ... 864 ... YESTERDAY.0 Removing input archive files ... done -Archive files older than 2 days being removed ... - NOW-5-old.-2H.MM.0 NOW-5-old.-2H.MM.index NOW-5-old.-2H.MM.meta NOW-5-old.-1H.MM.0 - NOW-5-old.-1H.MM.index NOW-5-old.-1H.MM.meta NOW-4-old.-2H.MM.0 NOW-4-old.-2H.MM.index - NOW-4-old.-2H.MM.meta NOW-4-old.-1H.MM.0 NOW-4-old.-1H.MM.index NOW-4-old.-1H.MM.meta - NOW-3-old.-2H.MM.0 NOW-3-old.-2H.MM.index NOW-3-old.-2H.MM.meta NOW-3-old.-1H.MM.0 - NOW-3-old.-1H.MM.index NOW-3-old.-1H.MM.meta NOW-5.-2H.MM.0 - NOW-5.-2H.MM.index NOW-5.-2H.MM.meta NOW-5.-1H.MM.0 - NOW-5.-1H.MM.index NOW-5.-1H.MM.meta NOW-4.-2H.MM.0 - NOW-4.-2H.MM.index NOW-4.-2H.MM.meta NOW-4.-1H.MM.0 - NOW-4.-1H.MM.index NOW-4.-1H.MM.meta NOW-3.-2H.MM.0 - NOW-3.-2H.MM.index NOW-3.-2H.MM.meta NOW-3.-1H.MM.0 - NOW-3.-1H.MM.index NOW-3.-1H.MM.meta -NOW-2-old.-1H.MM.0 -NOW-2-old.-1H.MM.index -NOW-2-old.-1H.MM.meta -NOW-2-old.-2H.MM.0 -NOW-2-old.-2H.MM.index -NOW-2-old.-2H.MM.meta -NOW-2.-1H.MM.0 -NOW-2.-1H.MM.index -NOW-2.-1H.MM.meta -NOW-2.-2H.MM.0 -NOW-2.-2H.MM.index -NOW-2.-2H.MM.meta -YESTERDAY-old.-1H.MM.0 -YESTERDAY-old.-1H.MM.index -YESTERDAY-old.-1H.MM.meta -YESTERDAY-old.-2H.MM.0 -YESTERDAY-old.-2H.MM.index -YESTERDAY-old.-2H.MM.meta -YESTERDAY.-1H.MM.0 -YESTERDAY.-1H.MM.index -YESTERDAY.-1H.MM.meta -YESTERDAY.-2H.MM.0 -YESTERDAY.-2H.MM.index -YESTERDAY.-2H.MM.meta +Archive files older than 3 days being removed ... + NOW-5.03.10.0 NOW-5.03.10.index NOW-5.03.10.meta NOW-5.05.10.0 + NOW-5.05.10.index NOW-5.05.10.meta NOW-4.03.10.0 NOW-4.03.10.index + NOW-4.03.10.meta NOW-4.05.10.0 NOW-4.05.10.index NOW-4.05.10.meta +NOW-2.03.10.0 +NOW-2.03.10.index +NOW-2.03.10.meta +NOW-2.05.10.0 +NOW-2.05.10.index +NOW-2.05.10.meta +NOW-3.03.10.0 +NOW-3.03.10.index +NOW-3.03.10.meta +NOW-3.05.10.0 +NOW-3.05.10.index +NOW-3.05.10.meta +TODAY.03.10.0 +TODAY.03.10.index +TODAY.03.10.meta +TODAY.05.10.0 +TODAY.05.10.index +TODAY.05.10.meta YESTERDAY.0 YESTERDAY.index YESTERDAY.meta @@ -149,27 +124,26 @@ sed PMLOGGER.DAILY: Error: no pmlogger instance running for host "HOST" [TMP/control:2] ... logging for host "HOST" unchanged -Archive files older than 1 days being removed ... - NOW-2-old.-2H.MM.0 NOW-2-old.-2H.MM.index NOW-2-old.-2H.MM.meta NOW-2-old.-1H.MM.0 - NOW-2-old.-1H.MM.index NOW-2-old.-1H.MM.meta NOW-2.-2H.MM.0 - NOW-2.-2H.MM.index NOW-2.-2H.MM.meta NOW-2.-1H.MM.0 - NOW-2.-1H.MM.index NOW-2.-1H.MM.meta -YESTERDAY-old.-1H.MM.0 -YESTERDAY-old.-1H.MM.index -YESTERDAY-old.-1H.MM.meta -YESTERDAY-old.-2H.MM.0 -YESTERDAY-old.-2H.MM.index -YESTERDAY-old.-2H.MM.meta -YESTERDAY.-1H.MM.0 -YESTERDAY.-1H.MM.index -YESTERDAY.-1H.MM.meta -YESTERDAY.-2H.MM.0 -YESTERDAY.-2H.MM.index -YESTERDAY.-2H.MM.meta -YESTERDAY.0 -YESTERDAY.index -YESTERDAY.meta +PMLOGGER.DAILY: Warning: no archives found to merge +Archive files older than 2 days being removed ... + NOW-3.03.10.0 NOW-3.03.10.index NOW-3.03.10.meta NOW-3.05.10.0 + NOW-3.05.10.index NOW-3.05.10.meta +NOW-2.03.10.0 +NOW-2.03.10.index +NOW-2.03.10.meta +NOW-2.05.10.0 +NOW-2.05.10.index +NOW-2.05.10.meta +TODAY.03.10.0 +TODAY.03.10.index +TODAY.03.10.meta +TODAY.05.10.0 +TODAY.05.10.index +TODAY.05.10.meta control +save.0 +save.index +save.meta sed === [verbose] error, output already exists === @@ -181,21 +155,18 @@ PMLOGGER.DAILY: Error: no pmlogger instance running for host "HOST" PMLOGGER.DAILY: Warning: output archive (YESTERDAY) already exists [TMP/control:2] ... skip log merging, culling and compressing for host "HOST" -YESTERDAY-old.-1H.MM.0 -YESTERDAY-old.-1H.MM.index -YESTERDAY-old.-1H.MM.meta -YESTERDAY-old.-2H.MM.0 -YESTERDAY-old.-2H.MM.index -YESTERDAY-old.-2H.MM.meta -YESTERDAY-old.18.19.0 -YESTERDAY-old.18.19.index -YESTERDAY-old.18.19.meta -YESTERDAY.-1H.MM.0 -YESTERDAY.-1H.MM.index -YESTERDAY.-1H.MM.meta -YESTERDAY.-2H.MM.0 -YESTERDAY.-2H.MM.index -YESTERDAY.-2H.MM.meta +NOW-2.03.10.0 +NOW-2.03.10.index +NOW-2.03.10.meta +NOW-2.05.10.0 +NOW-2.05.10.index +NOW-2.05.10.meta +TODAY.03.10.0 +TODAY.03.10.index +TODAY.03.10.meta +TODAY.05.10.0 +TODAY.05.10.index +TODAY.05.10.meta YESTERDAY.0 YESTERDAY.17.18.0 YESTERDAY.17.18.index diff --git a/532 b/532 index f75b501..0584bc1 100755 --- a/532 +++ b/532 @@ -2,8 +2,7 @@ # PCP QA Test No. 532 # $Revision: 1.16 $ # pmlogger_check with *-xx.* archives and .meta files in subdirs -# (based on 530) -# +# (loosely based on 530 and 338) # # Copyright (c) 1995-2002 Silicon Graphics, Inc. All Rights Reserved. # @@ -11,21 +10,6 @@ owner=kenmcd seq=`basename $0` -# Because pmlogger_daily is way too clever to use the same criteria -# for deciding which logs to merge and which to cull, we have to have -# QA output in 2 different times of the day - -case "`date +%H`" in -00) - # just after midnight - rm -f $seq.out && ln $seq.out.00-01 $seq.out || exit 1 - ;; -*) - # any other time - rm -f $seq.out && ln $seq.out.01-24 $seq.out || exit 1 - ;; -esac - echo "QA output created by $seq" # check for new version ... may need -o to force "old style" behaviour @@ -81,18 +65,31 @@ do for hour in 1 4 16 do [ $day -gt 0 -a $hour -gt 1 ] && continue - - namea=`pmdate -${day}d -${hour}H $fmt.%H.%M` - nameb=`pmdate -${day}d -${hour}H -1M $fmt.%H.%M` - namec=`pmdate -${day}d -${hour}H -2M $fmt.%H.%M` + # aiming for the time to be hour:10, hour:20 and hour:30 on + # each day, even if that is ahead of NOW for today + # + now_hr=`pmdate %H` + hr_set=`expr $hour - $now_hr` + [ "$hr_set" -ge 0 ] && hr_set="+$hr_set" + now_min=`pmdate %M` + min_set10=`expr 10 - $now_min` + [ "$min_set10" -ge 0 ] && min_set10="+$min_set10" + min_set20=`expr 20 - $now_min` + [ "$min_set20" -ge 0 ] && min_set20="+$min_set20" + min_set30=`expr 30 - $now_min` + [ "$min_set30" -ge 0 ] && min_set30="+$min_set30" + namea=`pmdate -${day}d ${hr_set}H ${min_set10}M $fmt.%H.%M` + nameb=`pmdate -${day}d ${hr_set}H ${min_set20}M $fmt.%H.%M` + namec=`pmdate -${day}d ${hr_set}H ${min_set30}M $fmt.%H.%M` + yesterday=`pmdate -1d ${hr_set}H ${min_set10}M $fmt` stampa=`echo $namea | sed -e 's/\.//g'` stampb=`echo $nameb | sed -e 's/\.//g'` stampc=`echo $namec | sed -e 's/\.//g'` nameb=${namea}-00 namec=${namea}-01 - echo "s/$namea/NOW-$day.-${hour}H.MM/g" >>$tmp/sed - echo "s/$nameb/NOW-$day.-${hour}H.MM-00/g" >>$tmp/sed - echo "s/$namec/NOW-$day.-${hour}H.MM-01/g" >>$tmp/sed + echo "s/$namea/NOW-$day.${hour}H.MM/g" >>$tmp/sed + echo "s/$nameb/NOW-$day.${hour}H.MM-00/g" >>$tmp/sed + echo "s/$namec/NOW-$day.${hour}H.MM-01/g" >>$tmp/sed for ext in 0 index meta do @@ -109,12 +106,13 @@ do done done -( cd $tmp; tar cf - *.0 *.meta *.index ) | ( cd $tmp/foo; tar xpf - ) +( cd $tmp; tar cf - *.0 *.meta *.index 2>/dev/null ) \ +| ( cd $tmp/foo; tar xpf - 2>/dev/null ) echo "s/NOW-0/TODAY/g" >>$tmp/sed echo "s/NOW-1/YESTERDAY/g" >>$tmp/sed echo "s/`pmdate %Y%m%d`/TODAY/g" >>$tmp/sed -echo "s/`pmdate -1d %Y%m%d`/YESTERDAY/g" >>$tmp/sed +echo "s/$yesterday/YESTERDAY/g" >>$tmp/sed echo >$tmp/control '$version=1.1 LOCALHOSTNAME n n '"$tmp"' ./config' diff --git a/564 b/564 index 48fa090..41895df 100755 --- a/564 +++ b/564 @@ -58,6 +58,8 @@ cp $tmp.ctl $tmp.relaydir/$FIVEDAYS.meta cp $tmp.ctl $tmp.relaydir/$FIVEDAYS.index cp $tmp.ctl $tmp.relaydir/mumble cp $tmp.ctl $tmp.relaydir/mumble.0 +STAMP=`pmdate -5d %Y%m%d`0010 +touch -t "$STAMP" $tmp.*/$FIVEDAYS.* $tmp.relaydir/mumble* # some archives for five months ago... FIVEMONTHS=`pmdate -5m %Y%m%d.0` @@ -67,6 +69,8 @@ cp $tmp.ctl $tmp.distdir/$FIVEMONTHS.index cp $tmp.ctl $tmp.relaydir/$FIVEMONTHS.0 cp $tmp.ctl $tmp.relaydir/$FIVEMONTHS.meta cp $tmp.ctl $tmp.relaydir/$FIVEMONTHS.index +STAMP=`pmdate -5m %Y%m%d`0010 +touch -t "$STAMP" $tmp.*/$FIVEMONTHS.* # real QA test starts here pmlogger_daily -V -N -x 3 -X bzip2 -c $tmp.ctl | sed \ @@ -74,6 +78,7 @@ pmlogger_daily -V -N -x 3 -X bzip2 -c $tmp.ctl | sed \ -e "s/$FIVEDAYS/FIVEDAYS/g" \ -e "s/$FIVEMONTHS/FIVEMONTHS/g" \ -e "s/$$/PID/g" \ + -e '/Warning: no archives found to merge/d' \ -e "s,/var/tmp,TMP,g" \ -e "s,/tmp,TMP,g" \ diff --git a/check b/check index d612233..4e8ba0c 100755 --- a/check +++ b/check @@ -429,6 +429,7 @@ do continue elif [ ! -f $seq ] then + echo " [not run, missing]" notrun="$notrun $seq" continue else diff --git a/common.filter b/common.filter index 030823c..6eba363 100644 --- a/common.filter +++ b/common.filter @@ -435,6 +435,7 @@ _filter_pcp_start() -e "s;$PCP_RC_DIR/pcp;\$PCP_RC_DIR/pcp;" \ -e "s;$PCP_BINADM_DIR/pmcd;\$PCP_BINADM_DIR/pmcd;" \ -e "s;$PCP_PMCDCONF_PATH;\$PCP_PMCDCONF_PATH;" \ + -e "s;$PCP_PMLOGGERCONTROL_PATH;\$PCP_PMLOGGERCONTROL_PATH;" \ -e "s;$PCP_VAR_DIR/;\$PCP_VAR_DIR/;" \ -e "s;/usr/etc/pmcd;\$PCP_BINADM_DIR/pmcd;" \ -e '/^Starting PCP$/d' \ diff --git a/group b/group index 5ca7870..f1e2144 100644 --- a/group +++ b/group @@ -910,3 +910,4 @@ dmf pcpqa 1234 pmchart local sgi 1236 #935639 pmchart local sgi 1237 pmchart pmgadgets local pmview sgi +338 logutil local pmlogextract local oss diff --git a/owner b/owner index 8dae782..8916e24 100644 --- a/owner +++ b/owner @@ -665,3 +665,4 @@ 1234 pcpqa 1236 pcpqa 1237 pcpqa +338 pcpqa diff --git a/show-me b/show-me index d79b2a0..a1fbd91 100755 --- a/show-me +++ b/show-me @@ -73,10 +73,10 @@ then # local and really me QA_DIR_DEFAULT=`pwd` else - QA_DIR_DEFAULT=`ssh $QA_USER@$QA_HOST /bin/ls -d isms/pcp[1-9]*/qa" 2>/dev/null | grep '^isms/' | tail -1` + QA_DIR_DEFAULT=`ssh $QA_USER@$QA_HOST "/bin/ls -d isms/pcp[1-9]*/qa" 2>/dev/null | grep '^isms/' | tail -1` fi else - QA_DIR_DEFAULT=`ssh $QA_USER@$QA_HOST /bin/ls -d isms/pcp[1-9]*/qa" 2>/dev/null | grep '^isms/' | tail -1` + QA_DIR_DEFAULT=`ssh $QA_USER@$QA_HOST "/bin/ls -d isms/pcp[1-9]*/qa" 2>/dev/null | grep '^isms/' | tail -1` fi if [ "$QA_DIR_DEFAULT" = "" ] then --=-eLWO3st/5zBNp7GM3PDX Content-Disposition: attachment; filename=338 Content-Type: application/x-shellscript; name=338 Content-Transfer-Encoding: 7bit #! /bin/sh # PCP QA Test No. 338 # $Revision: 1.1 $ # # pmlogger_daily ... do you work at all? # # NEW VERSION for YYYYDDMM style naming convention # # Copyright (c) 2008 Silicon Graphics, Inc. All Rights Reserved. # creator owner=kenj seq=`basename $0` echo "QA output created by $seq" # check for new version ... may need -o to force "old style" behaviour # if pmlogger_daily -\? 2>&1 | grep '^.*-o.*(old style)' >/dev/null then oflag='-o' else oflag='' fi # get standard environment, filters and checks . ./common.product . ./common.filter . ./common.check _filter() { tee -a $seq.full \ | $PCP_AWK_PROG ' /^Archive files older than/ { eatlines=1; longone=" "; print; next; } eatlines { if ( match ($0, "^ ") ) { for ( i=1; i <=NF; i++ ) { longone = sprintf ("%s %s",longone,$i); } } else { print longone | "fmt -76"; close ("fmt -76"); longone = ""; eatlines = 0; print; } next; } END { if ( eatlines ) { print longone | "fmt -76"; close ("fmt -76"); } } { print }' \ | sed -f $tmp/sed \ | sed \ -e "s;$tmp;TMP;g" \ -e "s/`hostname`/HOST/g" \ -e "s/`hostname | sed -e 's/\..*//'`/HOST/g" \ -e '/\/var\/adm\/pcplog\/NOTICES/d' \ | $PCP_AWK_PROG ' /-rw-r--r--/ { print $1,"...",$5,"...",$9; next } { print }' \ | _filter_cron_scripts } tmp=/tmp/$$ here=`pwd` sudo=$here/sudo status=1 # failure is the default! trap "cd /; rm -rf $tmp; $sudo mv $PCP_LOG_DIR/NOTICES.$$ $PCP_LOG_DIR/NOTICES; exit \$status" 0 1 2 3 15 $sudo rm -rf $tmp # real QA test starts here # take explicit control of the umask # umask 022 $sudo mv $PCP_LOG_DIR/NOTICES $PCP_LOG_DIR/NOTICES.$$ rm -f $seq.full mkdir $tmp # real QA test starts here unset verbose for loop in 1 2 do echo >$tmp/sed for day in 0 1 2 3 4 5 do fmt=%Y%m%d # aiming for the time to be 03:10 and 05:10 on each day, # even if that is ahead of NOW for today # now_hr=`pmdate %H` hra=`expr 3 - $now_hr` [ "$hra" -ge 0 ] && hra="+$hra" hrb=`expr 5 - $now_hr` [ "$hrb" -ge 0 ] && hrb="+$hrb" now_min=`pmdate %M` min=`expr 10 - $now_min` [ "$min" -ge 0 ] && min="+$min" namea=`pmdate -${day}d ${hra}H ${min}M $fmt.%H.%M` nameb=`pmdate -${day}d ${hrb}H ${min}M $fmt.%H.%M` yesterday=`pmdate -1d ${hrb}H ${min}M $fmt` stampa=`echo $namea | sed -e 's/\.//g'` stampb=`echo $nameb | sed -e 's/\.//g'` echo "s/$namea/NOW-$day.03.10/g" >>$tmp/sed echo "s/$nameb/NOW-$day.05.10/g" >>$tmp/sed for ext in 0 index meta do cp src-oss/arch_a.$ext $tmp/$namea.$ext touch -t $stampa $tmp/$namea.$ext cp src-oss/arch_b.$ext $tmp/$nameb.$ext touch -t $stampb $tmp/$nameb.$ext done done ls -lt $tmp >>$seq.full echo "s/NOW-0/TODAY/g" >>$tmp/sed echo "s/NOW-1/YESTERDAY/g" >>$tmp/sed echo "s/`pmdate %Y%m%d`/TODAY/g" >>$tmp/sed echo "s/$yesterday/YESTERDAY/g" >>$tmp/sed echo >$tmp/control '$version=1.1 LOCALHOSTNAME n n '"$tmp"' ./config' echo echo "=== [${verbose+verbose}] regular show ===" echo "=== [${verbose+verbose}] regular show ===" >>$seq.full pmlogger_daily $oflag -c $tmp/control -k 3 $verbose 2>&1 \ | _filter ls $tmp | sed -f $tmp/sed | LC_COLLATE=POSIX sort echo echo "=== [${verbose+verbose}] nothing to do but cull some more ===" echo "=== [${verbose+verbose}] nothing to do but cull some more ===" >>$seq.full for ext in 0 index meta do mv $tmp/$yesterday.$ext $tmp/save.$ext done pmlogger_daily $oflag -c $tmp/control -k 2 $verbose 2>&1 \ | _filter ls $tmp | sed -f $tmp/sed | LC_COLLATE=POSIX sort echo echo "=== [${verbose+verbose}] error, output already exists ===" echo "=== [${verbose+verbose}] error, output already exists ===" >>$seq.full for ext in 0 index meta do cp src-oss/arch_a.$ext $tmp/$yesterday.17.18.$ext mv $tmp/save.$ext $tmp/$yesterday.$ext done pmlogger_daily $oflag -c $tmp/control -k 2 $verbose 2>&1 \ | _filter ls $tmp | sed -f $tmp/sed | LC_COLLATE=POSIX sort verbose=-V rm -f $tmp/* done # all done status=0 exit --=-eLWO3st/5zBNp7GM3PDX Content-Disposition: attachment; filename=338.out Content-Type: text/plain; name=338.out; charset=UTF-8 Content-Transfer-Encoding: 7bit QA output created by 338 === [] regular show === PMLOGGER.DAILY: Error: no pmlogger instance running for host "HOST" [TMP/control:2] ... logging for host "HOST" unchanged NOW-2.03.10.0 NOW-2.03.10.index NOW-2.03.10.meta NOW-2.05.10.0 NOW-2.05.10.index NOW-2.05.10.meta NOW-3.03.10.0 NOW-3.03.10.index NOW-3.03.10.meta NOW-3.05.10.0 NOW-3.05.10.index NOW-3.05.10.meta TODAY.03.10.0 TODAY.03.10.index TODAY.03.10.meta TODAY.05.10.0 TODAY.05.10.index TODAY.05.10.meta YESTERDAY.0 YESTERDAY.index YESTERDAY.meta control sed === [] nothing to do but cull some more === PMLOGGER.DAILY: Error: no pmlogger instance running for host "HOST" [TMP/control:2] ... logging for host "HOST" unchanged NOW-2.03.10.0 NOW-2.03.10.index NOW-2.03.10.meta NOW-2.05.10.0 NOW-2.05.10.index NOW-2.05.10.meta TODAY.03.10.0 TODAY.03.10.index TODAY.03.10.meta TODAY.05.10.0 TODAY.05.10.index TODAY.05.10.meta control save.0 save.index save.meta sed === [] error, output already exists === PMLOGGER.DAILY: Error: no pmlogger instance running for host "HOST" [TMP/control:2] ... logging for host "HOST" unchanged PMLOGGER.DAILY: Warning: output archive (YESTERDAY) already exists [TMP/control:2] ... skip log merging, culling and compressing for host "HOST" NOW-2.03.10.0 NOW-2.03.10.index NOW-2.03.10.meta NOW-2.05.10.0 NOW-2.05.10.index NOW-2.05.10.meta TODAY.03.10.0 TODAY.03.10.index TODAY.03.10.meta TODAY.05.10.0 TODAY.05.10.index TODAY.05.10.meta YESTERDAY.0 YESTERDAY.17.18.0 YESTERDAY.17.18.index YESTERDAY.17.18.meta YESTERDAY.index YESTERDAY.meta control sed === [verbose] regular show === === daily maintenance of PCP archives for host HOST === PMLOGGER.DAILY: Error: no pmlogger instance running for host "HOST" [TMP/control:2] ... logging for host "HOST" unchanged Input archives to be merged: YESTERDAY.03.10 YESTERDAY.05.10 Output archive files: -rw-r--r-- ... 455 ... YESTERDAY.meta -rw-r--r-- ... 192 ... YESTERDAY.index -rw-r--r-- ... 864 ... YESTERDAY.0 Removing input archive files ... done Archive files older than 3 days being removed ... NOW-5.03.10.0 NOW-5.03.10.index NOW-5.03.10.meta NOW-5.05.10.0 NOW-5.05.10.index NOW-5.05.10.meta NOW-4.03.10.0 NOW-4.03.10.index NOW-4.03.10.meta NOW-4.05.10.0 NOW-4.05.10.index NOW-4.05.10.meta NOW-2.03.10.0 NOW-2.03.10.index NOW-2.03.10.meta NOW-2.05.10.0 NOW-2.05.10.index NOW-2.05.10.meta NOW-3.03.10.0 NOW-3.03.10.index NOW-3.03.10.meta NOW-3.05.10.0 NOW-3.05.10.index NOW-3.05.10.meta TODAY.03.10.0 TODAY.03.10.index TODAY.03.10.meta TODAY.05.10.0 TODAY.05.10.index TODAY.05.10.meta YESTERDAY.0 YESTERDAY.index YESTERDAY.meta control sed === [verbose] nothing to do but cull some more === === daily maintenance of PCP archives for host HOST === PMLOGGER.DAILY: Error: no pmlogger instance running for host "HOST" [TMP/control:2] ... logging for host "HOST" unchanged PMLOGGER.DAILY: Warning: no archives found to merge Archive files older than 2 days being removed ... NOW-3.03.10.0 NOW-3.03.10.index NOW-3.03.10.meta NOW-3.05.10.0 NOW-3.05.10.index NOW-3.05.10.meta NOW-2.03.10.0 NOW-2.03.10.index NOW-2.03.10.meta NOW-2.05.10.0 NOW-2.05.10.index NOW-2.05.10.meta TODAY.03.10.0 TODAY.03.10.index TODAY.03.10.meta TODAY.05.10.0 TODAY.05.10.index TODAY.05.10.meta control save.0 save.index save.meta sed === [verbose] error, output already exists === === daily maintenance of PCP archives for host HOST === PMLOGGER.DAILY: Error: no pmlogger instance running for host "HOST" [TMP/control:2] ... logging for host "HOST" unchanged PMLOGGER.DAILY: Warning: output archive (YESTERDAY) already exists [TMP/control:2] ... skip log merging, culling and compressing for host "HOST" NOW-2.03.10.0 NOW-2.03.10.index NOW-2.03.10.meta NOW-2.05.10.0 NOW-2.05.10.index NOW-2.05.10.meta TODAY.03.10.0 TODAY.03.10.index TODAY.03.10.meta TODAY.05.10.0 TODAY.05.10.index TODAY.05.10.meta YESTERDAY.0 YESTERDAY.17.18.0 YESTERDAY.17.18.index YESTERDAY.17.18.meta YESTERDAY.index YESTERDAY.meta control sed --=-eLWO3st/5zBNp7GM3PDX-- From kenj@internode.on.net Wed Nov 26 04:17:41 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=0.4 required=5.0 tests=J_CHICKENPOX_54, J_CHICKENPOX_64,J_CHICKENPOX_65,J_CHICKENPOX_75 autolearn=no version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAQAHeZv004240 for ; Wed, 26 Nov 2008 04:17:40 -0600 X-ASG-Debug-ID: 1227694657-19cf00c30000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from mail.internode.on.net (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id A9AE4160485C for ; Wed, 26 Nov 2008 02:17:37 -0800 (PST) Received: from mail.internode.on.net (bld-mail05.adl2.internode.on.net [203.16.214.69]) by cuda.sgi.com with ESMTP id udzVBBdbpjggpmSe for ; Wed, 26 Nov 2008 02:17:37 -0800 (PST) Received: from [192.168.1.100] (unverified [118.208.222.17]) by mail.internode.on.net (SurgeMail 3.8f2) with ESMTP id 35110515-1927428 for ; Wed, 26 Nov 2008 19:31:08 +1030 (CDT) X-ASG-Orig-Subj: pmlogger_daily daylight saving change fix and other goodies Subject: pmlogger_daily daylight saving change fix and other goodies From: Ken McDonell Reply-To: kenj@internode.on.net To: pcp@oss.sgi.com Content-Type: multipart/mixed; boundary="=-XbRCv+xSU5UQKFpeKs1r" Date: Wed, 26 Nov 2008 20:00:49 +1100 Message-Id: <1227690049.2540.13.camel@bozo> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 X-Barracuda-Connect: bld-mail05.adl2.internode.on.net[203.16.214.69] X-Barracuda-Start-Time: 1227694659 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -2.02 X-Barracuda-Spam-Status: No, SCORE=-2.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests= X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11194 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- --=-XbRCv+xSU5UQKFpeKs1r Content-Type: text/plain Content-Transfer-Encoding: 7bit I don't have access to oss.sgi.com yet, so here is another patch that fixes the pmlogger_daily problem with daylight saving transitions. The problem does lie with pmdate (and this is not fixable), and there were two related but different problems that happened when daylight savings started and ended. This patch has been checked multiple times with QA $ check -g logutil -x remote -x sgi on a couple of different machines (once you have the associated pcpcqa patch applied). For the brave soul applying the patch, here is some take text ... man/man1/pmlogger_daily.1 src/pmlogctl/pmlogger_daily.sh Change the handling of time in the past to not use pmdate, and algorithmic changes to avoid the problems associated with the transitions at both ends of daylight savings periods. Culling and compressing are now based on the mtime of the archives. Independent of when the script is run, today's archives are NEVER candidates for merging. Resurrected the lost Aztec mode in which a superior script will merge all archives (not just yesterday's archives) by default. Added -o to support the old Spanish mode where only yesterday's archives are merge candidates. --=-XbRCv+xSU5UQKFpeKs1r Content-Disposition: attachment; filename=pcp.patch Content-Type: text/x-patch; name=pcp.patch; charset=UTF-8 Content-Transfer-Encoding: 7bit diff --git a/man/man1/pmlogger_daily.1 b/man/man1/pmlogger_daily.1 index 4c6e44d..b0e6c06 100644 --- a/man/man1/pmlogger_daily.1 +++ b/man/man1/pmlogger_daily.1 @@ -37,7 +37,7 @@ \f3pmlogger_merge\f1 \- administration of Performance Co-Pilot archive log files .SH SYNOPSIS .B $PCP_BINADM_DIR/pmlogger_daily -[\f3\-NV\f1] +[\f3\-NoV\f1] [\f3\-c\f1 \f2control\f1] [\f3\-k\f1 \f2discard\f1] [\f3\-m\f1 \f2addresses\f1] @@ -405,6 +405,11 @@ in conjunction with .B \-V maximizes the diagnostic capabilities for debugging. .PP +By default all possible archives will be merged. The +.B \-o +option reinstates the old behaviour in which only yesterday's archives +will be considered as merge candidates. +.PP The .B \-T option provides a terser form of output for diff --git a/src/pmlogctl/pmlogger_daily.sh b/src/pmlogctl/pmlogger_daily.sh index 7502e60..33ab025 100644 --- a/src/pmlogctl/pmlogger_daily.sh +++ b/src/pmlogctl/pmlogger_daily.sh @@ -102,6 +102,8 @@ Options: -k discard remove archives after "discard" days -m addresses send daily NOTICES entries to email addresses -N show-me mode, no operations performed + -o (old style) merge logs only from yesterday + [default is to merge all possible logs before today] -V verbose output -x compress compress archive data files after "compress" days -X program use program for archive data file compression @@ -117,8 +119,9 @@ SHOWME=false VERBOSE=false VERY_VERBOSE=false MYARGS="" +OFLAG=false -while getopts c:k:m:Ns:Vx:X:Y:? c +while getopts c:k:m:Nos:Vx:X:Y:? c do case $c in @@ -138,6 +141,8 @@ do N) SHOWME=true MYARGS="$MYARGS -N" ;; + o) OFLAG=true + ;; s) ROLLNOTICES="$OPTARG" check=`echo "$ROLLNOTICES" | sed -e 's/[0-9]//g'` if [ ! -z "$check" ] @@ -188,12 +193,6 @@ fi # LOGNAME=`date "+%Y%m%d.%H.%M"` -# each summarized log is named yyyymmdd using yesterday's date -# previous day's logs are named yymmdd (old format) or -# yyyymmdd (new year 2000 format) -# -SUMMARY_LOGNAME=`pmdate -1d %Y%m%d` - _error() { _report Error "$1" @@ -217,57 +216,21 @@ _unlock() echo >$tmp.lock } -# filter for pmlogger archive files in working directory - -# pass in the number of days to skip over (backwards) from today +# filter file names to leave those that look like PCP archives +# managed by pmlogger_check and pmlogger_daily, namely they begin +# with a datestamp # -# pv:821339 too many sed commands for IRIX ... split into groups -# of at most 200 days +# need to handle both the year 2000 and the old name formats, and +# possible ./ prefix (from find .) # -_date_filter() +_filter_filename() { - # start with all files whose names match the patterns used by - # the PCP archive log management scripts ... this list may be - # reduced by the sed filtering later on - # - # need to handle both the year 2000 and the old name formats - # - ls | sed -n >$tmp.in \ + sed -n \ + -e 's/^\.\///' \ -e '/^[12][0-9][0-9][0-9][0-1][0-9][0-3][0-9][-.]/p' \ -e '/^[0-9][0-9][0-1][0-9][0-3][0-9][-.]/p' - - i=0 - while [ $i -le $1 ] - do - dmax=`expr $i + 200` - [ $dmax -gt $1 ] && dmax=$1 - echo "/^[12][0-9][0-9][0-9][0-1][0-9][0-3][0-9][-.]/{" >$tmp.sed1 - echo "/^[0-9][0-9][0-1][0-9][0-3][0-9][-.]/{" >$tmp.sed2 - while [ $i -le $dmax ] - do - x=`pmdate -${i}d %Y%m%d` - echo "/^$x\./d" >>$tmp.sed1 - echo "/^$x-[0-9][0-9]\./d" >>$tmp.sed1 - x=`pmdate -${i}d %y%m%d` - echo "/^$x\./d" >>$tmp.sed2 - echo "/^$x-[0-9][0-9]\./d" >>$tmp.sed2 - i=`expr $i + 1` - done - echo "p" >>$tmp.sed1 - echo "}" >>$tmp.sed1 - echo "p" >>$tmp.sed2 - echo "}" >>$tmp.sed2 - cat $tmp.sed2 >>$tmp.sed1 - - # cull file names with matching dates, keep other file names - # - sed -n -f $tmp.sed1 <$tmp.in >$tmp.tmp - mv $tmp.tmp $tmp.in - done - - cat $tmp.in } - # mails out any entries for the previous 24hrs from the PCP notices file # if [ ! -z "$MAILME" ] @@ -567,10 +530,18 @@ s/^\([A-Za-z][A-Za-z0-9_]*\)=/export \1; \1=/p fqdn=`pmhostname $host` for log in $PCP_TMP_DIR/pmlogger/[0-9]* do - [ "$log" = "[0-9]*" ] && continue + case "$log" + in + *[0-9]*) + # no pmlogger running according to $PCP_TMP_DIR/pmlogger, + # nothing to be done! + _unlock + continue + ;; + esac $VERY_VERBOSE && $PCP_ECHO_PROG $PCP_ECHO_N "... try $log: ""$PCP_ECHO_C" match=`sed -e '3s/\/[0-9][0-9][0-9][0-9][0-9.]*$//' $log \ - | $PCP_AWK_PROG ' + | $PCP_AWK_PROG ' BEGIN { m = 0 } NR == 2 && $1 == "'$fqdn'" { m = 1; next } NR == 2 && "'$fqdn'" == "'$host'" && @@ -622,55 +593,112 @@ END { print m }'` touch $tmp.err fi fi + $VERBOSE && echo - # concatenate yesterday's archive logs + # Merge archive logs. # - # note: we need to handle duplicate-breaking forms like + # Will work for new style YYYYMMDD.HH.MM[-NN] archives and old style + # YYMMDD.HH.MM[-NN] archives. + # Note: we need to handle duplicate-breaking forms like # YYYYMMDD.HH.MM-seq# (even though pmlogger_merge already picks most # of these up) in case the base YYYYMMDD.HH.MM archive is for some # reason missing here # - $VERBOSE && echo + # Assume if .meta file is present then other archive components are + # also present (if not the case it is a serious process botch, and + # pmlogmerge will fail below) + # + # Find all candidate input archives, remove any that contain today's + # date and group the remainder by date. + # + TODAY=`date +%Y%m%d` + + find *.meta \ + \( -name "*.[0-2][0-9].[0-5][0-9].meta" \ + -o -name "*.[0-2][0-9].[0-5][0-9]-[0-9][0-9].meta" \ + \) \ + -print \ + | sed \ + -e "/^$TODAY\./d" \ + -e 's/\.meta//' \ + | sort -n \ + | $PCP_AWK_PROG ' + { if (lastdate != "" && match($1, "^" lastdate "\\.") == 1) { + # same date as previous one + inlist = inlist " " $1 + next + } + else { + # different date as previous one + if (inlist != "") print lastdate,inlist + inlist = $1 + lastdate = $1 + sub(/\..*/, "", lastdate) + } + } +END { if (inlist != "") print lastdate,inlist }' >$tmp.list + + if $OFLAG + then + # -o option, preserve the old semantics, and only process the + # previous day's archives ... aim for a time close to midday + # yesterday and report that date + # + now_hr=`pmdate %H` + hr=`expr 12 + $now_hr` + grep "^[0-9]*`pmdate -${hr}H %y%m%d` " $tmp.list >$tmp.tmp + mv $tmp.tmp $tmp.list + fi - WANT_LOG=`find *.meta \ - \( -name "*.[0-2][0-9].[0-5][0-9].meta" \ - -o -name "*.[0-2][0-9].[0-5][0-9]-[0-9][0-9].meta" \ - \) \ - -newer $tmp.merge \ - -print \ - | sed \ - -e "/$LOGNAME/d" \ - -e 's/\.meta//' \ - -e 's/^\.\///'` - - if [ ! -z "$WANT_LOG" ] + rm -f $tmp.skip + if [ ! -s $tmp.list ] then - if [ -f $SUMMARY_LOGNAME.0 -o -f $SUMMARY_LOGNAME.index -o -f $SUMMARY_LOGNAME.meta ] + if $VERBOSE then - echo "$prog: Warning: output archive ($SUMMARY_LOGNAME) already exists" - echo "[$CONTROL:$line] ... skip log merging, culling and compressing for host \"$host\"" - echo "Note: Possibly a daylight saving change caused a large date jump?" - continue - else - if $SHOWME + echo "$prog: Warning: no archives found to merge" + $VERY_VERBOSE && ls -l + fi + else + cat $tmp.list \ + | while read outfile inlist + do + if [ -f $outfile.0 -o -f $outfile.index -o -f $outfile.meta ] then - echo "+ pmlogger_merge$MYARGS -f $WANT_LOG $SUMMARY_LOGNAME" + echo "$prog: Warning: output archive ($outfile) already exists" + echo "[$CONTROL:$line] ... skip log merging, culling and compressing for host \"$host\"" + touch $tmp.skip + break else - if pmlogger_merge$MYARGS -f $WANT_LOG $SUMMARY_LOGNAME + if $SHOWME then - : + echo "+ pmlogger_merge$MYARGS -f $inlist $outfile" else - _error "problems executing pmlogger_merge for host \"$host\"" + if pmlogger_merge$MYARGS -f $inlist $outfile + then + : + else + _error "problems executing pmlogger_merge for host \"$host\"" + fi fi fi - fi + done + fi + + if [ -f $tmp.skip ] + then + # this is sufficiently serious that we don't want to remove + # the lock file, so problems are not compounded the next time + # the script is run + continue fi # and cull old archives # if [ X"$CULLAFTER" != X"forever" ] then - _date_filter $CULLAFTER >$tmp.list + find . -type f -mtime +$CULLAFTER \ + | _filter_filename \ + | sort >$tmp.list if [ -s $tmp.list ] then if $VERBOSE @@ -693,7 +721,10 @@ END { print m }'` if [ ! -z "$COMPRESSAFTER" ] then - _date_filter $COMPRESSAFTER | egrep -v "$COMPRESSREGEX" >$tmp.list + find . -type f -mtime +$COMPRESSAFTER \ + | _filter_filename \ + | egrep -v "$COMPRESSREGEX" \ + | sort >$tmp.list if [ -s $tmp.list ] then if $VERBOSE --=-XbRCv+xSU5UQKFpeKs1r-- From nscott@aconex.com Wed Nov 26 16:44:45 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=-1.8 required=5.0 tests=AWL,BAYES_00 autolearn=ham version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAQMijvM016502 for ; Wed, 26 Nov 2008 16:44:45 -0600 X-ASG-Debug-ID: 1227739483-708b01550000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from postoffice2.aconex.com (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 757F216092FF for ; Wed, 26 Nov 2008 14:44:43 -0800 (PST) Received: from postoffice2.aconex.com (mail.aconex.com [203.89.202.182]) by cuda.sgi.com with ESMTP id E7cPF0GnUiBqu7zN for ; Wed, 26 Nov 2008 14:44:43 -0800 (PST) Received: from postoffice.aconex.com (localhost [127.0.0.1]) by postoffice2.aconex.com (Spam Firewall) with ESMTP id DD3336732FF; Thu, 27 Nov 2008 09:44:39 +1100 (EST) Received: from postoffice.aconex.com (postoffice.yarra.acx [192.168.102.1]) by postoffice2.aconex.com with ESMTP id xLXYbqI1TeryuGPB; Thu, 27 Nov 2008 09:44:39 +1100 (EST) Received: from [192.168.5.24] (melho0.aconex.com [203.89.192.141]) by postoffice.aconex.com (Postfix) with ESMTP id C12B292C2E2; Thu, 27 Nov 2008 09:44:39 +1100 (EST) X-ASG-Orig-Subj: Re: [pcp] pmlogger_daily daylight saving change fix and other goodies Subject: Re: [pcp] pmlogger_daily daylight saving change fix and other goodies From: Nathan Scott To: kenj@internode.on.net Cc: pcp@oss.sgi.com In-Reply-To: <1227690049.2540.13.camel@bozo> References: <1227690049.2540.13.camel@bozo> Content-Type: text/plain Date: Thu, 27 Nov 2008 09:42:19 +1100 Message-Id: <1227739339.4624.41.camel@verge.scott.net.au> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: mail.aconex.com[203.89.202.182] X-Barracuda-Start-Time: 1227739484 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0002 1.0000 -2.0199 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -2.02 X-Barracuda-Spam-Status: No, SCORE=-2.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests= X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11243 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- On Wed, 2008-11-26 at 20:00 +1100, Ken McDonell wrote: > I don't have access to oss.sgi.com yet, so here is another patch that > fixes the pmlogger_daily problem with daylight saving transitions. Great - I look forward to the next daylight savings switch with much less trepidation! One question about the changes ... @@ -567,10 +530,18 @@ s/^\([A-Za-z][A-Za-z0-9_]*\)=/export \1; \1=/p fqdn=`pmhostname $host` for log in $PCP_TMP_DIR/pmlogger/[0-9]* do - [ "$log" = "[0-9]*" ] && continue + case "$log" + in + *[0-9]*) + # no pmlogger running according to $PCP_TMP_DIR/pmlogger, + # nothing to be done! + _unlock This _unlock and continue pair looks a bit dubious. We're going to "continue" the inner loop here not the outer one, which means we'd potentially prematurely _unlock, wouldn't we? + continue + ;; + esac cheers. -- Nathan From kenj@internode.on.net Wed Nov 26 22:42:13 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=-2.2 required=5.0 tests=BAYES_00,J_CHICKENPOX_54, J_CHICKENPOX_64,J_CHICKENPOX_65,J_CHICKENPOX_75 autolearn=no version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAR4gCTj001129 for ; Wed, 26 Nov 2008 22:42:13 -0600 X-ASG-Debug-ID: 1227760929-4baf01390000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from mail.internode.on.net (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id D73521211EA9 for ; Wed, 26 Nov 2008 20:42:10 -0800 (PST) Received: from mail.internode.on.net (bld-mail09.adl2.internode.on.net [203.16.214.73]) by cuda.sgi.com with ESMTP id lyCBHCOS0EWL8MEb for ; Wed, 26 Nov 2008 20:42:10 -0800 (PST) Received: from [192.168.1.100] (unverified [118.208.222.17]) by mail.internode.on.net (SurgeMail 3.8f2) with ESMTP id 49465400-1927428 for multiple; Thu, 27 Nov 2008 15:12:07 +1030 (CDT) X-ASG-Orig-Subj: Re: [pcp] pmlogger_daily daylight saving change fix and other goodies Subject: Re: [pcp] pmlogger_daily daylight saving change fix and other goodies From: Ken McDonell Reply-To: kenj@internode.on.net To: Nathan Scott Cc: pcp@oss.sgi.com In-Reply-To: <1227739339.4624.41.camel@verge.scott.net.au> References: <1227690049.2540.13.camel@bozo> <1227739339.4624.41.camel@verge.scott.net.au> Content-Type: multipart/mixed; boundary="=-+5xxIsPfb92YmJVMLTqc" Date: Thu, 27 Nov 2008 15:41:47 +1100 Message-Id: <1227760907.6207.22.camel@bozo> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 X-Barracuda-Connect: bld-mail09.adl2.internode.on.net[203.16.214.73] X-Barracuda-Start-Time: 1227760931 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -2.02 X-Barracuda-Spam-Status: No, SCORE=-2.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests= X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11266 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- --=-+5xxIsPfb92YmJVMLTqc Content-Type: text/plain Content-Transfer-Encoding: 7bit Thanks Nathan. Yes the _unlock was a little enthusiastic. The continue is OK as there is no log files in /var/tmp/pcp/pmlogger to inspect and we should just fall into the code after this for loop as there are no more arguments for this for loop ... but break is probably clearer than continue, so I've made that change also. The rediffed patch is attached. On Thu, 2008-11-27 at 09:42 +1100, Nathan Scott wrote: > On Wed, 2008-11-26 at 20:00 +1100, Ken McDonell wrote: > > I don't have access to oss.sgi.com yet, so here is another patch that > > fixes the pmlogger_daily problem with daylight saving transitions. > > Great - I look forward to the next daylight savings switch with > much less trepidation! > > One question about the changes ... > > @@ -567,10 +530,18 @@ s/^\([A-Za-z][A-Za-z0-9_]*\)=/export \1; \1=/p > fqdn=`pmhostname $host` > for log in $PCP_TMP_DIR/pmlogger/[0-9]* > do > - [ "$log" = "[0-9]*" ] && continue > + case "$log" > + in > + *[0-9]*) > + # no pmlogger running according to > $PCP_TMP_DIR/pmlogger, > + # nothing to be done! > + _unlock > > This _unlock and continue pair looks a bit dubious. We're going > to "continue" the inner loop here not the outer one, which means > we'd potentially prematurely _unlock, wouldn't we? > > + continue > + ;; > + esac > > > cheers. > > -- > Nathan > --=-+5xxIsPfb92YmJVMLTqc Content-Disposition: attachment; filename=pcp.patch Content-Type: text/x-patch; name=pcp.patch; charset=utf-8 Content-Transfer-Encoding: 7bit diff --git a/man/man1/pmlogger_daily.1 b/man/man1/pmlogger_daily.1 index 4c6e44d..b0e6c06 100644 --- a/man/man1/pmlogger_daily.1 +++ b/man/man1/pmlogger_daily.1 @@ -37,7 +37,7 @@ \f3pmlogger_merge\f1 \- administration of Performance Co-Pilot archive log files .SH SYNOPSIS .B $PCP_BINADM_DIR/pmlogger_daily -[\f3\-NV\f1] +[\f3\-NoV\f1] [\f3\-c\f1 \f2control\f1] [\f3\-k\f1 \f2discard\f1] [\f3\-m\f1 \f2addresses\f1] @@ -405,6 +405,11 @@ in conjunction with .B \-V maximizes the diagnostic capabilities for debugging. .PP +By default all possible archives will be merged. The +.B \-o +option reinstates the old behaviour in which only yesterday's archives +will be considered as merge candidates. +.PP The .B \-T option provides a terser form of output for diff --git a/src/pmlogctl/pmlogger_daily.sh b/src/pmlogctl/pmlogger_daily.sh index 7502e60..1911bae 100644 --- a/src/pmlogctl/pmlogger_daily.sh +++ b/src/pmlogctl/pmlogger_daily.sh @@ -102,6 +102,8 @@ Options: -k discard remove archives after "discard" days -m addresses send daily NOTICES entries to email addresses -N show-me mode, no operations performed + -o (old style) merge logs only from yesterday + [default is to merge all possible logs before today] -V verbose output -x compress compress archive data files after "compress" days -X program use program for archive data file compression @@ -117,8 +119,9 @@ SHOWME=false VERBOSE=false VERY_VERBOSE=false MYARGS="" +OFLAG=false -while getopts c:k:m:Ns:Vx:X:Y:? c +while getopts c:k:m:Nos:Vx:X:Y:? c do case $c in @@ -138,6 +141,8 @@ do N) SHOWME=true MYARGS="$MYARGS -N" ;; + o) OFLAG=true + ;; s) ROLLNOTICES="$OPTARG" check=`echo "$ROLLNOTICES" | sed -e 's/[0-9]//g'` if [ ! -z "$check" ] @@ -188,12 +193,6 @@ fi # LOGNAME=`date "+%Y%m%d.%H.%M"` -# each summarized log is named yyyymmdd using yesterday's date -# previous day's logs are named yymmdd (old format) or -# yyyymmdd (new year 2000 format) -# -SUMMARY_LOGNAME=`pmdate -1d %Y%m%d` - _error() { _report Error "$1" @@ -217,57 +216,21 @@ _unlock() echo >$tmp.lock } -# filter for pmlogger archive files in working directory - -# pass in the number of days to skip over (backwards) from today +# filter file names to leave those that look like PCP archives +# managed by pmlogger_check and pmlogger_daily, namely they begin +# with a datestamp # -# pv:821339 too many sed commands for IRIX ... split into groups -# of at most 200 days +# need to handle both the year 2000 and the old name formats, and +# possible ./ prefix (from find .) # -_date_filter() +_filter_filename() { - # start with all files whose names match the patterns used by - # the PCP archive log management scripts ... this list may be - # reduced by the sed filtering later on - # - # need to handle both the year 2000 and the old name formats - # - ls | sed -n >$tmp.in \ + sed -n \ + -e 's/^\.\///' \ -e '/^[12][0-9][0-9][0-9][0-1][0-9][0-3][0-9][-.]/p' \ -e '/^[0-9][0-9][0-1][0-9][0-3][0-9][-.]/p' - - i=0 - while [ $i -le $1 ] - do - dmax=`expr $i + 200` - [ $dmax -gt $1 ] && dmax=$1 - echo "/^[12][0-9][0-9][0-9][0-1][0-9][0-3][0-9][-.]/{" >$tmp.sed1 - echo "/^[0-9][0-9][0-1][0-9][0-3][0-9][-.]/{" >$tmp.sed2 - while [ $i -le $dmax ] - do - x=`pmdate -${i}d %Y%m%d` - echo "/^$x\./d" >>$tmp.sed1 - echo "/^$x-[0-9][0-9]\./d" >>$tmp.sed1 - x=`pmdate -${i}d %y%m%d` - echo "/^$x\./d" >>$tmp.sed2 - echo "/^$x-[0-9][0-9]\./d" >>$tmp.sed2 - i=`expr $i + 1` - done - echo "p" >>$tmp.sed1 - echo "}" >>$tmp.sed1 - echo "p" >>$tmp.sed2 - echo "}" >>$tmp.sed2 - cat $tmp.sed2 >>$tmp.sed1 - - # cull file names with matching dates, keep other file names - # - sed -n -f $tmp.sed1 <$tmp.in >$tmp.tmp - mv $tmp.tmp $tmp.in - done - - cat $tmp.in } - # mails out any entries for the previous 24hrs from the PCP notices file # if [ ! -z "$MAILME" ] @@ -567,10 +530,17 @@ s/^\([A-Za-z][A-Za-z0-9_]*\)=/export \1; \1=/p fqdn=`pmhostname $host` for log in $PCP_TMP_DIR/pmlogger/[0-9]* do - [ "$log" = "[0-9]*" ] && continue + case "$log" + in + *[0-9]*) + # no pmlogger running according to $PCP_TMP_DIR/pmlogger + # + break + ;; + esac $VERY_VERBOSE && $PCP_ECHO_PROG $PCP_ECHO_N "... try $log: ""$PCP_ECHO_C" match=`sed -e '3s/\/[0-9][0-9][0-9][0-9][0-9.]*$//' $log \ - | $PCP_AWK_PROG ' + | $PCP_AWK_PROG ' BEGIN { m = 0 } NR == 2 && $1 == "'$fqdn'" { m = 1; next } NR == 2 && "'$fqdn'" == "'$host'" && @@ -622,55 +592,112 @@ END { print m }'` touch $tmp.err fi fi + $VERBOSE && echo - # concatenate yesterday's archive logs + # Merge archive logs. # - # note: we need to handle duplicate-breaking forms like + # Will work for new style YYYYMMDD.HH.MM[-NN] archives and old style + # YYMMDD.HH.MM[-NN] archives. + # Note: we need to handle duplicate-breaking forms like # YYYYMMDD.HH.MM-seq# (even though pmlogger_merge already picks most # of these up) in case the base YYYYMMDD.HH.MM archive is for some # reason missing here # - $VERBOSE && echo + # Assume if .meta file is present then other archive components are + # also present (if not the case it is a serious process botch, and + # pmlogmerge will fail below) + # + # Find all candidate input archives, remove any that contain today's + # date and group the remainder by date. + # + TODAY=`date +%Y%m%d` + + find *.meta \ + \( -name "*.[0-2][0-9].[0-5][0-9].meta" \ + -o -name "*.[0-2][0-9].[0-5][0-9]-[0-9][0-9].meta" \ + \) \ + -print \ + | sed \ + -e "/^$TODAY\./d" \ + -e 's/\.meta//' \ + | sort -n \ + | $PCP_AWK_PROG ' + { if (lastdate != "" && match($1, "^" lastdate "\\.") == 1) { + # same date as previous one + inlist = inlist " " $1 + next + } + else { + # different date as previous one + if (inlist != "") print lastdate,inlist + inlist = $1 + lastdate = $1 + sub(/\..*/, "", lastdate) + } + } +END { if (inlist != "") print lastdate,inlist }' >$tmp.list + + if $OFLAG + then + # -o option, preserve the old semantics, and only process the + # previous day's archives ... aim for a time close to midday + # yesterday and report that date + # + now_hr=`pmdate %H` + hr=`expr 12 + $now_hr` + grep "^[0-9]*`pmdate -${hr}H %y%m%d` " $tmp.list >$tmp.tmp + mv $tmp.tmp $tmp.list + fi - WANT_LOG=`find *.meta \ - \( -name "*.[0-2][0-9].[0-5][0-9].meta" \ - -o -name "*.[0-2][0-9].[0-5][0-9]-[0-9][0-9].meta" \ - \) \ - -newer $tmp.merge \ - -print \ - | sed \ - -e "/$LOGNAME/d" \ - -e 's/\.meta//' \ - -e 's/^\.\///'` - - if [ ! -z "$WANT_LOG" ] + rm -f $tmp.skip + if [ ! -s $tmp.list ] then - if [ -f $SUMMARY_LOGNAME.0 -o -f $SUMMARY_LOGNAME.index -o -f $SUMMARY_LOGNAME.meta ] + if $VERBOSE then - echo "$prog: Warning: output archive ($SUMMARY_LOGNAME) already exists" - echo "[$CONTROL:$line] ... skip log merging, culling and compressing for host \"$host\"" - echo "Note: Possibly a daylight saving change caused a large date jump?" - continue - else - if $SHOWME + echo "$prog: Warning: no archives found to merge" + $VERY_VERBOSE && ls -l + fi + else + cat $tmp.list \ + | while read outfile inlist + do + if [ -f $outfile.0 -o -f $outfile.index -o -f $outfile.meta ] then - echo "+ pmlogger_merge$MYARGS -f $WANT_LOG $SUMMARY_LOGNAME" + echo "$prog: Warning: output archive ($outfile) already exists" + echo "[$CONTROL:$line] ... skip log merging, culling and compressing for host \"$host\"" + touch $tmp.skip + break else - if pmlogger_merge$MYARGS -f $WANT_LOG $SUMMARY_LOGNAME + if $SHOWME then - : + echo "+ pmlogger_merge$MYARGS -f $inlist $outfile" else - _error "problems executing pmlogger_merge for host \"$host\"" + if pmlogger_merge$MYARGS -f $inlist $outfile + then + : + else + _error "problems executing pmlogger_merge for host \"$host\"" + fi fi fi - fi + done + fi + + if [ -f $tmp.skip ] + then + # this is sufficiently serious that we don't want to remove + # the lock file, so problems are not compounded the next time + # the script is run + continue fi # and cull old archives # if [ X"$CULLAFTER" != X"forever" ] then - _date_filter $CULLAFTER >$tmp.list + find . -type f -mtime +$CULLAFTER \ + | _filter_filename \ + | sort >$tmp.list if [ -s $tmp.list ] then if $VERBOSE @@ -693,7 +720,10 @@ END { print m }'` if [ ! -z "$COMPRESSAFTER" ] then - _date_filter $COMPRESSAFTER | egrep -v "$COMPRESSREGEX" >$tmp.list + find . -type f -mtime +$COMPRESSAFTER \ + | _filter_filename \ + | egrep -v "$COMPRESSREGEX" \ + | sort >$tmp.list if [ -s $tmp.list ] then if $VERBOSE --=-+5xxIsPfb92YmJVMLTqc-- From nscott@aconex.com Thu Nov 27 01:08:49 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=-1.8 required=5.0 tests=AWL,BAYES_00 autolearn=ham version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mAR78m6V014480 for ; Thu, 27 Nov 2008 01:08:49 -0600 X-ASG-Debug-ID: 1227769726-56c400930000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from postoffice2.aconex.com (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id A132B16073EC for ; Wed, 26 Nov 2008 23:08:47 -0800 (PST) Received: from postoffice2.aconex.com (mail.aconex.com [203.89.202.182]) by cuda.sgi.com with ESMTP id Wc9jWo4tlaDIBF9B for ; Wed, 26 Nov 2008 23:08:47 -0800 (PST) Received: from postoffice.aconex.com (localhost [127.0.0.1]) by postoffice2.aconex.com (Spam Firewall) with ESMTP id 93CFF674D7B for ; Thu, 27 Nov 2008 16:30:34 +1100 (EST) Received: from postoffice.aconex.com (postoffice.yarra.acx [192.168.102.1]) by postoffice2.aconex.com with ESMTP id yCyURImrNDbZTC2x for ; Thu, 27 Nov 2008 16:30:34 +1100 (EST) Received: from [192.168.5.24] (melho0.aconex.com [203.89.192.141]) by postoffice.aconex.com (Postfix) with ESMTP id 71DEF92C2F9 for ; Thu, 27 Nov 2008 16:30:34 +1100 (EST) X-ASG-Orig-Subj: pcp updates Subject: pcp updates From: Nathan Scott To: pcp@oss.sgi.com Content-Type: text/plain Date: Thu, 27 Nov 2008 16:28:13 +1100 Message-Id: <1227763693.4624.81.camel@verge.scott.net.au> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: mail.aconex.com[203.89.202.182] X-Barracuda-Start-Time: 1227769727 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -2.02 X-Barracuda-Spam-Status: No, SCORE=-2.02 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests= X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11276 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- Changes committed to git://oss.sgi.com/nathans/pcp.git man/man1/pmlogger_daily.1 | 7 + src/pmlogctl/pmlogger_daily.sh | 194 +++++++++++++++++++++++------------------ 2 files changed, 118 insertions(+), 83 deletions(-) commit 3514c44f4910c70067c7a1656841080b1b68d7e7 Merge: ba54f2c... c3223ff... Author: Nathan Scott Date: Thu Nov 27 16:22:23 2008 +1100 Merge branch 'dev' of git://oss.sgi.com/pcp/pcp into nathans commit ba54f2c1bbc327e90a57269580bf06a44fea36d3 Author: Ken McDonell Date: Thu Nov 27 16:15:14 2008 +1100 Fix the pmlogger_daily problem with daylight saving transitions. Change the handling of time in the past to not use pmdate, and algorithmic changes to avoid the problems associated with the transitions at both ends of daylight savings periods. Culling and compressing are now based on the mtime of the archives. Independent of when the script is run, today's archives are NEVER candidates for merging. Resurrected the lost Aztec mode in which a superior script will merge all archives (not just yesterday's archives) by default. Added -o to support the old Spanish mode where only yesterday's archives are merge candidates. From nscott@aconex.com Sun Nov 30 19:08:49 2008 X-Spam-Checker-Version: SpamAssassin 3.3.0-rupdated (updated) on oss.sgi.com X-Spam-Level: X-Spam-Status: No, score=-1.8 required=5.0 tests=AWL,BAYES_00,J_CHICKENPOX_66 autolearn=no version=3.3.0-rupdated Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id mB118miK002912 for ; Sun, 30 Nov 2008 19:08:49 -0600 X-ASG-Debug-ID: 1228093726-66c700c20000-S8gJnT X-Barracuda-URL: http://cuda.sgi.com:80/cgi-bin/mark.cgi Received: from postoffice2.aconex.com (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id 09ACE162C26E for ; Sun, 30 Nov 2008 17:08:47 -0800 (PST) Received: from postoffice2.aconex.com (mail.aconex.com [203.89.202.182]) by cuda.sgi.com with ESMTP id uwhatJSnpQKfzui4 for ; Sun, 30 Nov 2008 17:08:47 -0800 (PST) Received: from postoffice.aconex.com (localhost [127.0.0.1]) by postoffice2.aconex.com (Spam Firewall) with ESMTP id BA07944FF3A for ; Mon, 1 Dec 2008 11:56:18 +1100 (EST) Received: from postoffice.aconex.com (postoffice.yarra.acx [192.168.102.1]) by postoffice2.aconex.com with ESMTP id CUPzMVmnN3a7Rs6S for ; Mon, 01 Dec 2008 11:56:18 +1100 (EST) Received: from [192.168.5.24] (melho0.aconex.com [203.89.192.141]) by postoffice.aconex.com (Postfix) with ESMTP id 8161692C2E2 for ; Mon, 1 Dec 2008 11:56:18 +1100 (EST) X-ASG-Orig-Subj: pcp updates Subject: pcp updates From: Nathan Scott To: pcp@oss.sgi.com Content-Type: text/plain Date: Mon, 01 Dec 2008 11:53:58 +1100 Message-Id: <1228092838.8036.36.camel@verge.scott.net.au> Mime-Version: 1.0 X-Mailer: Evolution 2.22.3.1 Content-Transfer-Encoding: 7bit X-Barracuda-Connect: mail.aconex.com[203.89.202.182] X-Barracuda-Start-Time: 1228093728 X-Barracuda-Bayes: INNOCENT GLOBAL 0.0000 1.0000 -2.0210 X-Barracuda-Virus-Scanned: by cuda.sgi.com at sgi.com X-Barracuda-Spam-Score: -1.52 X-Barracuda-Spam-Status: No, SCORE=-1.52 using per-user scores of TAG_LEVEL=2.0 QUARANTINE_LEVEL=1000.0 KILL_LEVEL=2.1 tests=BSF_RULE7568M X-Barracuda-Spam-Report: Code version 3.2, rules version 3.2.1.11607 Rule breakdown below pts rule name description ---- ---------------------- -------------------------------------------------- 0.50 BSF_RULE7568M Custom Rule 7568M Changes committed to git://oss.sgi.com/nathans/pcpqa.git 094 | 15 +-- 094.out | 45 --------- 151 | 99 +++++++++++++-------- 151.out | 195 ++++++++++++++++++++++++++++++++++++++++++ 151.out.00-01 | 237 ---------------------------------------------------- 151.out.01-24 | 237 ---------------------------------------------------- 172 | 2 183 | 7 - 183.out.linux | 6 - 191 | 3 202 | 19 +++- 282 | 24 ++--- 282.out.3 | 19 ++-- 338 | 166 ++++++++++++++++++++++++++++++++++++ 338.out | 178 +++++++++++++++++++++++++++++++++++++++ 347 | 2 398 | 5 - 530 | 90 ++++++++----------- 530.out | 177 ++++++++++++++++++++++++++++++++++++++ 530.out.00-01 | 137 ------------------------------ 530.out.01-02 | 172 ------------------------------------- 530.out.02-24 | 206 --------------------------------------------- 532 | 50 +++++----- 532.out | 150 ++++++++++++++++++++++++++++++++ 532.out.00-01 | 164 ----------------------------------- 532.out.01-24 | 154 --------------------------------- 564 | 5 + 580 | 1 check | 1 common.filter | 1 group | 1 owner | 1 show-me | 4 src-oss/permfetch.c | 27 ++--- 34 files changed, 1074 insertions(+), 1526 deletions(-) commit 1a030abd301b8b435b75b9a84ae9294da35b497c Author: Ken McDonell Date: Sun Nov 30 20:24:26 2008 +1100 Filter changes because init is /sbin/init sometimes. commit e6c43298320af13a9c6b0dda115e25769d1916dc Author: Ken McDonell Date: Sun Nov 30 20:19:42 2008 +1100 Modify pmcp init script filter to avoid additional pmcd_wait diag messages. commit 00126318752f2ce66d3aaf5e7fd0b2d81d8bb026 Author: Ken McDonell Date: Sun Nov 30 20:10:33 2008 +1100 Fine tune filter to support gmake or make in the install step. commit 54d3b79cbe074a9ca60d23a5aa1bd3837dd41d11 Author: Ken McDonell Date: Sun Nov 30 20:03:57 2008 +1100 Remove blank line from filtered kill output @ end. commit 3f23a4d5893e5e605644c1c2608cf0955ed2381c Author: Ken McDonell Date: Sun Nov 30 20:00:48 2008 +1100 Tighten up the egrep pattern to avoid easter eggs like ... unix 2 [ ACC ] STREAM LISTENING 15892 /tmp/orbit-kenj/linc-1861-0-b42000fb2ff4 commit a39332475ce39d4a92d902bc0344a65f79245c06 Author: Ken McDonell Date: Sun Nov 30 15:27:24 2008 +1100 Rework 094 to use pmval rather than pmkstat (which has been retired). commit bed2eb3380eaeb79d789ce795037c53f3ba4523d Author: Ken McDonell Date: Sun Nov 30 15:05:03 2008 +1100 Fix malloc use-after-free error in test application. Exposed by more recent versions of gcc. commit cc97c9a94de61aaa4cc43a320e21c06f242d155d Author: Ken McDonell Date: Sun Nov 30 12:53:32 2008 +1100 New qualified output for 530 after pmlogger_daily changes. commit 2d0865a4a9e6f0beeb5edf38050f81f231423f47 Author: Ken McDonell Date: Sun Nov 30 12:51:26 2008 +1100 Dual of 530 to separate handling of new (YYYY) format from old (YY) format tests in the wake of the pmlogger_daily changes to accommodate DST change windows. commit 3e2b94735e513c012749402280aa92f8e0ccb4b6 Author: Ken McDonell Date: Sun Nov 30 12:49:31 2008 +1100 Changes to accommodate pmlogger_daily rework to fix DST change bugs by removing dependence on pmdate. Subtle changes to -k and -x options to pmlogger_daily (using file mtime and name of file, rather than pmdate and name of file) required some rework of the associated QA tests. commit 7c00546f4ca3896695a50455382895e12285a295 Author: Ken McDonell Date: Sun Nov 30 12:44:20 2008 +1100 Changes to accommodate pmlogger_daily rework to fix DST change bugs by removing dependence on pmdate. Subtle changes to -k and -x options to pmlogger_daily (using file mtime and name of file, rather than pmdate and name of file) required some rework of the associated QA tests. commit ddc16af48c79e1caa1f51f0ba27b02a37e90053e Author: Ken McDonell Date: Sun Nov 30 12:13:48 2008 +1100 Residual changes from parameterization of pmlogger_* control file path. commit 8934eb0ba97a304e1a818ffe90b7e49959ba1130 Author: Ken McDonell Date: Sun Nov 30 11:54:50 2008 +1100 Improve interactive reporting when tests are not run because the file is named in "group" but does not exist (happens for people outside sgi for all tests that are not open sourced). commit fdfedec88ff38e4dd15df40b5922815281535add Author: Ken McDonell Date: Sun Nov 30 10:24:30 2008 +1100 Fix quoting botch from previous checkin.