From owner-pcp@oss.sgi.com Tue Jul 3 03:15:35 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f63AFZ316977 for pcp-outgoing; Tue, 3 Jul 2001 03:15:35 -0700 Received: from mailgw1.netvision.net.il (mailgw1.netvision.net.il [194.90.1.14]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f63AFWV16968 for ; Tue, 3 Jul 2001 03:15:32 -0700 Received: from mail.exanet.co.il ([212.143.73.99]) by mailgw1.netvision.net.il (8.9.3/8.9.3) with ESMTP id NAA22500 for ; Tue, 3 Jul 2001 13:15:28 +0300 (IDT) Received: from cat (fw1.exanet.co.il [212.143.73.98]) by mail.exanet.co.il (8.11.0/8.11.0) with SMTP id f63ACng16100 for ; Tue, 3 Jul 2001 13:12:51 +0300 Reply-To: From: "gilly" To: Subject: multiple PMDA's with one root Date: Tue, 3 Jul 2001 13:14:04 +0200 Message-ID: <015101c103b1$459916c0$2a04000a@exanet> MIME-Version: 1.0 Content-Type: text/plain; charset="windows-1255" Content-Transfer-Encoding: 7bit X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook CWS, Build 9.0.2416 (9.0.2910.0) Importance: Normal X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400 Sender: owner-pcp@oss.sgi.com Precedence: bulk Is it possible to create different PMDA's with one root in the PMNS? For example - if I use different PMDA's for different linux services (PMDA's like - linux_nfs, linux_disk, linux_kernel, etc...), but I want all these PMDA's to be set under one root named linux in the PMNS. If so, where should I define this 'linux' root ('cause the different PMDA's don't know of each other)? From owner-pcp@oss.sgi.com Tue Jul 3 03:28:52 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f63ASqJ18902 for pcp-outgoing; Tue, 3 Jul 2001 03:28:52 -0700 Received: from pneumatic-tube.sgi.com (pneumatic-tube.sgi.com [204.94.214.22]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f63ASqV18899 for ; Tue, 3 Jul 2001 03:28:52 -0700 Received: from larry.melbourne.sgi.com (larry.melbourne.sgi.com [134.14.52.130]) by pneumatic-tube.sgi.com (980327.SGI.8.8.8-aspam/980310.SGI-aspam) via SMTP id DAA03462 for ; Tue, 3 Jul 2001 03:26:08 -0700 (PDT) mail_from (markgw@sgi.com) Received: from sherman.melbourne.sgi.com (sherman.melbourne.sgi.com [134.14.55.175]) by larry.melbourne.sgi.com (950413.SGI.8.6.12/950213.SGI.AUTOCF) via ESMTP id UAA25352; Tue, 3 Jul 2001 20:27:33 +1000 Date: Tue, 3 Jul 2001 20:27:33 +1000 (EST) From: Mark Goodwin X-Sender: markgw@sherman.melbourne.sgi.com To: gilly cc: pcp@oss.sgi.com Subject: Re: multiple PMDA's with one root In-Reply-To: <015101c103b1$459916c0$2a04000a@exanet> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pcp@oss.sgi.com Precedence: bulk On Tue, 3 Jul 2001, gilly wrote: > Is it possible to create different PMDA's with one root in the PMNS? > For example - if I use different PMDA's for different linux services (PMDA's > like - linux_nfs, linux_disk, linux_kernel, etc...), but I want all these > PMDA's to be set under one root named linux in the PMNS. > If so, where should I define this 'linux' root ('cause the different PMDA's > don't know of each other)? > Just duplicate the non-leaf nodes in the PMNS for each PMDA. Each PMDA "Install" script calls pmnsadd(1) and/or pmnsmerge(1) to merge a subtree into an existing PMNS. -- Mark From owner-pcp@oss.sgi.com Tue Jul 3 09:28:59 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f63GSx326444 for pcp-outgoing; Tue, 3 Jul 2001 09:28:59 -0700 Received: from mailgw2.netvision.net.il (mailgw2.netvision.net.il [194.90.1.9]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f63GSuV26441 for ; Tue, 3 Jul 2001 09:28:56 -0700 Received: from mail.exanet.co.il ([212.143.73.99]) by mailgw2.netvision.net.il (8.9.3/8.9.3) with ESMTP id TAA12662 for ; Tue, 3 Jul 2001 19:30:58 +0300 (IDT) Received: from cat (fw1.exanet.co.il [212.143.73.98]) by mail.exanet.co.il (8.11.0/8.11.0) with SMTP id f63GQ5g19724 for ; Tue, 3 Jul 2001 19:26:05 +0300 Reply-To: From: "gilly" To: Subject: RE: multiple PMDA's with one root Date: Tue, 3 Jul 2001 19:27:27 +0200 Message-ID: <016501c103e5$6edd5e40$2a04000a@exanet> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook CWS, Build 9.0.2416 (9.0.2910.0) Importance: Normal X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400 In-Reply-To: Sender: owner-pcp@oss.sgi.com Precedence: bulk thanks for the quick reply, but I encounter another problem concerning the same issue - I have a different ID for each PMDA and I don't know where should I define an ID for the root. does the PMDA_PMID for each metric in each PMDA should change (right now it's PMDA_PMID(0,x) since the tree consists of 2 levels only). I pass the compilation phase, but while running "Install" script I get PM_ERR_NAME (it doesn't recognize the PMDA name). -----Original Message----- From: Mark Goodwin [mailto:markgw@sgi.com] Sent: Tuesday, July 03, 2001 12:28 PM To: gilly Cc: pcp@oss.sgi.com Subject: Re: multiple PMDA's with one root On Tue, 3 Jul 2001, gilly wrote: > Is it possible to create different PMDA's with one root in the PMNS? > For example - if I use different PMDA's for different linux services (PMDA's > like - linux_nfs, linux_disk, linux_kernel, etc...), but I want all these > PMDA's to be set under one root named linux in the PMNS. > If so, where should I define this 'linux' root ('cause the different PMDA's > don't know of each other)? > Just duplicate the non-leaf nodes in the PMNS for each PMDA. Each PMDA "Install" script calls pmnsadd(1) and/or pmnsmerge(1) to merge a subtree into an existing PMNS. -- Mark From owner-pcp@oss.sgi.com Wed Jul 4 04:15:45 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f64BFjF12733 for pcp-outgoing; Wed, 4 Jul 2001 04:15:45 -0700 Received: from mailgw3.netvision.net.il (mailgw3.netvision.net.il [194.90.1.11]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f64BFdV12730 for ; Wed, 4 Jul 2001 04:15:42 -0700 Received: from mail.exanet.co.il ([212.143.73.99]) by mailgw3.netvision.net.il (8.9.3/8.9.3) with ESMTP id MAA06116 for ; Wed, 4 Jul 2001 12:59:28 +0300 (IDT) Received: from cat (fw1.exanet.co.il [212.143.73.98]) by mail.exanet.co.il (8.11.0/8.11.0) with SMTP id f649wTg26960 for ; Wed, 4 Jul 2001 12:58:29 +0300 Reply-To: From: "gilly" To: Subject: multiple PMDA's with one root - continue Date: Wed, 4 Jul 2001 12:59:52 +0200 Message-ID: <019001c10478$73cebdc0$2a04000a@exanet> MIME-Version: 1.0 Content-Type: text/plain; charset="windows-1255" Content-Transfer-Encoding: 7bit X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook CWS, Build 9.0.2416 (9.0.2910.0) Importance: Normal X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400 Sender: owner-pcp@oss.sgi.com Precedence: bulk I've succeeded implementing different PMDA's under one root. However, there is a problem with pmdaInstall in 'pmdaproc.sh' script (called from the 'Install' script under each PMDA). In order to check that the agent is running OK this function calls pminfo over all metrics in the pmns by "pmns_name" where "pmns_name=$iam" (the line is: " for __n in $pmns_name ... pminfo -n $NAMESPACE -f $__n" However, in my case, I should call 'pminfo -f' with the root name as a prefix before $iam. Regarding the previous example: if the root is 'linux' and two PMDA's - 'nfs' & 'memory'. The script sends to 'pminfo -f' the metrics 'nfs' & 'memory', instead of the metrics 'linux.nfs' & 'linux.memory'. I just want to be sure 'm not missing anything here and what is the importance of this check under '_install'. thanks, gilly From owner-pcp@oss.sgi.com Wed Jul 4 17:39:08 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f650d8s03168 for pcp-outgoing; Wed, 4 Jul 2001 17:39:08 -0700 Received: from deliverator.sgi.com (deliverator.sgi.com [204.94.214.10]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f650d7V03165 for ; Wed, 4 Jul 2001 17:39:07 -0700 Received: from rattle.melbourne.sgi.com (rattle.melbourne.sgi.com [134.14.55.145]) by deliverator.sgi.com (980309.SGI.8.8.8-aspam-6.2/980310.SGI-aspam) via ESMTP id RAA02313 for ; Wed, 4 Jul 2001 17:38:59 -0700 (PDT) mail_from (kenmcd@melbourne.sgi.com) Received: from localhost (kenmcd@localhost) by rattle.melbourne.sgi.com (SGI-8.9.3/8.9.3) with ESMTP id KAA72527; Thu, 5 Jul 2001 10:37:47 +1000 (AEST) X-Authentication-Warning: rattle.melbourne.sgi.com: kenmcd owned process doing -bs Date: Thu, 5 Jul 2001 10:37:47 +1000 From: Ken McDonell Reply-To: To: gilly cc: Subject: Re: multiple PMDA's with one root In-Reply-To: <015101c103b1$459916c0$2a04000a@exanet> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pcp@oss.sgi.com Precedence: bulk On Tue, 3 Jul 2001, gilly wrote: > Is it possible to create different PMDA's with one root in the PMNS? > For example - if I use different PMDA's for different linux services (PMDA's > like - linux_nfs, linux_disk, linux_kernel, etc...), but I want all these > PMDA's to be set under one root named linux in the PMNS. I think this is both possible and may be not possible (it depends _exactly_ on what you want to do). It is possible in the sense that the PMNS is a tree, that provides a taxonomy from names to PMIDs and it is the domain (top-level) part of the PMID that identifies the PMDA that knows how to answer questions about a particular metric. So, this is a valid PMNS $ pminfo -m -n root linux.nfs.one PMID: 80.0.0 linux.nfs.two PMID: 80.0.1 linux.disk.three PMID: 81.0.0 linux.disk.four PMID: 81.0.1 linux.kernel.mumble PMID: 82.0.0 linux.kernel.fumble PMID: 82.0.1 linux.kernel.stumble PMID: 82.0.2 The will continue to work as long as the metrics provided by each PMDA end up in non-overlapping parts of the PMNS. This is true in the example above. But it is not true here (where I have promoted the linux.kernel.* metrics to be linux.* metrics): $ pminfo -m -n root linux.nfs.one PMID: 80.0.0 linux.nfs.two PMID: 80.0.1 linux.disk.three PMID: 81.0.0 linux.disk.four PMID: 81.0.1 linux.mumble PMID: 82.0.0 linux.fumble PMID: 82.0.1 linux.stumble PMID: 82.0.2 The problem here is that if you remove the PMDA providing mumble, fumble and stumble (i.e. domain 82) then you must remove all metrics starting at their common prefix, i.e. at linux ... this will also remove the linux.nfs and linux.disk metrics which is a bad thing. It turns out that adding a PMDA starts by culling any of its old names (using the same name prefix procedure, since the old names cannot be correctly enumerated if they belong to an old PMDA that has been replaced by a new version of the same PMDA), and then installing the new names. So _either_ Installing or Removing the domain 82 PMDA would also remove the names for the metrics from domain 80 and 81 in the PMNS above ... clearly this is a BAD thing. Note this is a function of the way we update the PMNS in the scripts that sit behind the PMDA Install and Remove scripts, rather than the design of the PMNS per se. > If so, where should I define this 'linux' root ('cause the different PMDA's > don't know of each other)? Something like this in /var/pcp/pmns/root ... root { ... linux ... } linux { nfs disk kernel } linux.nfs { ... } etc. - From owner-pcp@oss.sgi.com Wed Jul 4 17:39:13 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f650dDZ03185 for pcp-outgoing; Wed, 4 Jul 2001 17:39:13 -0700 Received: from deliverator.sgi.com (deliverator.sgi.com [204.94.214.10]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f650dDV03182 for ; Wed, 4 Jul 2001 17:39:13 -0700 Received: from rattle.melbourne.sgi.com (rattle.melbourne.sgi.com [134.14.55.145]) by deliverator.sgi.com (980309.SGI.8.8.8-aspam-6.2/980310.SGI-aspam) via ESMTP id RAA02317 for ; Wed, 4 Jul 2001 17:39:05 -0700 (PDT) mail_from (kenmcd@melbourne.sgi.com) Received: from localhost (kenmcd@localhost) by rattle.melbourne.sgi.com (SGI-8.9.3/8.9.3) with ESMTP id KAA02197; Thu, 5 Jul 2001 10:37:55 +1000 (AEST) X-Authentication-Warning: rattle.melbourne.sgi.com: kenmcd owned process doing -bs Date: Thu, 5 Jul 2001 10:37:54 +1000 From: Ken McDonell Reply-To: To: gilly cc: Subject: RE: multiple PMDA's with one root In-Reply-To: <016501c103e5$6edd5e40$2a04000a@exanet> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pcp@oss.sgi.com Precedence: bulk On Tue, 3 Jul 2001, gilly wrote: > thanks for the quick reply, but I encounter another problem concerning the > same issue - I have a different ID for each PMDA and I don't know where > should I define an ID for the root. does the PMDA_PMID for each metric in > each PMDA should change (right now it's PMDA_PMID(0,x) since the tree > consists of 2 levels only). > I pass the compilation phase, but while running "Install" script I get > PM_ERR_NAME (it doesn't recognize the PMDA name). You don't need a domain number for "linux". You need difference domain numbers for each PMDA providing linux.foo, linux.bar, etc. See my previous reply for more details. From owner-pcp@oss.sgi.com Wed Jul 4 17:39:22 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f650dMh03212 for pcp-outgoing; Wed, 4 Jul 2001 17:39:22 -0700 Received: from deliverator.sgi.com (deliverator.sgi.com [204.94.214.10]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f650dLV03209 for ; Wed, 4 Jul 2001 17:39:21 -0700 Received: from rattle.melbourne.sgi.com (rattle.melbourne.sgi.com [134.14.55.145]) by deliverator.sgi.com (980309.SGI.8.8.8-aspam-6.2/980310.SGI-aspam) via ESMTP id RAA02321 for ; Wed, 4 Jul 2001 17:39:13 -0700 (PDT) mail_from (kenmcd@melbourne.sgi.com) Received: from localhost (kenmcd@localhost) by rattle.melbourne.sgi.com (SGI-8.9.3/8.9.3) with ESMTP id KAA78078; Thu, 5 Jul 2001 10:38:03 +1000 (AEST) X-Authentication-Warning: rattle.melbourne.sgi.com: kenmcd owned process doing -bs Date: Thu, 5 Jul 2001 10:38:03 +1000 From: Ken McDonell Reply-To: To: gilly cc: Subject: Re: multiple PMDA's with one root - continue In-Reply-To: <019001c10478$73cebdc0$2a04000a@exanet> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pcp@oss.sgi.com Precedence: bulk On Wed, 4 Jul 2001, gilly wrote: > I've succeeded implementing different PMDA's under one root. However, there > is a problem with pmdaInstall in 'pmdaproc.sh' script (called from the > 'Install' script under each PMDA). In order to check that the agent is > running OK this function calls pminfo over all metrics in the pmns by > "pmns_name" where "pmns_name=$iam" (the line is: > " for __n in $pmns_name ... pminfo -n $NAMESPACE -f $__n" > However, in my case, I should call 'pminfo -f' with the root name as a > prefix before $iam. > Regarding the previous example: > if the root is 'linux' and two PMDA's - 'nfs' & 'memory'. The script sends > to 'pminfo -f' the metrics 'nfs' & 'memory', instead of the metrics > 'linux.nfs' & 'linux.memory'. > I just want to be sure 'm not missing anything here and what is the > importance of this check under '_install'. > thanks, gilly In your Install (and Remove) scripts, after setting iam, over-ride the default $pmns_name e.g. iam=linux-nfs # this is the PMDA name pmns_name=linux.nfs # this is the PMNS prefix for names this PMDA provides From owner-pcp@oss.sgi.com Thu Jul 5 10:12:05 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f65HC5L22574 for pcp-outgoing; Thu, 5 Jul 2001 10:12:05 -0700 Received: from rigel.cis.ksu.edu (root@rigel.cis.ksu.edu [129.130.10.65]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f65HC3V22571 for ; Thu, 5 Jul 2001 10:12:03 -0700 Received: from pollux.cis.ksu.edu (mst9696@pollux.cis.ksu.edu [129.130.10.33]) by rigel.cis.ksu.edu (8.9.1/8.9.1/000517) with ESMTP id MAA20980 for ; Thu, 5 Jul 2001 12:12:02 -0500 (CDT) Received: from localhost (mst9696@localhost) by pollux.cis.ksu.edu (8.9.1/8.9.1/000517) with ESMTP id MAA22106 for ; Thu, 5 Jul 2001 12:12:01 -0500 (CDT) X-Authentication-Warning: pollux.cis.ksu.edu: mst9696 owned process doing -bs Date: Thu, 5 Jul 2001 12:12:01 -0500 (CDT) From: Madhu Sudhan R Tera To: pcp@oss.sgi.com Subject: getting the currently running process metrics Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pcp@oss.sgi.com Precedence: bulk Hello all! I am new to this mailing list! Currently i am working on load balancing in beowulf cluster. The proposed procedure is to get the metrics for all the running processes from PCP and provide these metrics as input to the loadbalancing algorithm, which evaluates these metrics and decides which process to be migrated. But I dont find any facility to get the process metrics like CPU usage by each running process or memorey required by each process.pmem tool is not available in non-commercial versions of PCP,i guess! Please tell me how i can get the process metrics programmatically using pcp libraries.DO i have to write a PMDA or PMAPI?? I am not clear on this. Thank you, With regards, MadhuSudhan Reddy Tera -------------------------------------------------- |Department of Computing and Information Sciences, | |Kansas State University, | |Manhattan, KS 66502. | |ph no:785-537-3276 | -------------------------------------------------- From owner-pcp@oss.sgi.com Thu Jul 5 12:49:04 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f65Jn4D26658 for pcp-outgoing; Thu, 5 Jul 2001 12:49:04 -0700 Received: from sgi.com (sgi.SGI.COM [192.48.153.1]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f65Jn3V26655 for ; Thu, 5 Jul 2001 12:49:03 -0700 Received: from rattle.melbourne.sgi.com (rattle.melbourne.sgi.com [134.14.55.145]) by sgi.com (980327.SGI.8.8.8-aspam/980304.SGI-aspam: SGI does not authorize the use of its proprietary systems or networks for unsolicited or bulk email from the Internet.) via ESMTP id MAA08866 for ; Thu, 5 Jul 2001 12:48:57 -0700 (PDT) mail_from (kenmcd@melbourne.sgi.com) Received: from localhost (kenmcd@localhost) by rattle.melbourne.sgi.com (SGI-8.9.3/8.9.3) with ESMTP id FAA58565; Fri, 6 Jul 2001 05:47:37 +1000 (AEST) X-Authentication-Warning: rattle.melbourne.sgi.com: kenmcd owned process doing -bs Date: Fri, 6 Jul 2001 05:47:37 +1000 From: Ken McDonell Reply-To: To: Madhu Sudhan R Tera cc: Subject: Re: getting the currently running process metrics In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pcp@oss.sgi.com Precedence: bulk On Thu, 5 Jul 2001, Madhu Sudhan R Tera wrote: > Hello all! > I am new to this mailing list! Welcome. > Currently i am working on load balancing in beowulf cluster. > > The proposed procedure is to get the metrics for all the running > processes from PCP and provide these metrics as input to the loadbalancing > algorithm, which evaluates these metrics and decides which process to be > migrated. Seems like a reasonable fit with PCP's capabilities. > But I dont find any facility to get the process metrics like CPU usage by > each running process or memorey required by each process.pmem tool is not > available in non-commercial versions of PCP,i guess! The reason pmem is not there is that compared to IRIX, Linux does not export the same (nor as detailed) information about the address space of each process. But there are some stats ... check out pminfo -dT proc.memory Similarly the CPU usage stats are covered by pminfo -dT proc.psinfo.utime proc.psinfo.stime proc.psinfo.cutime \ proc.psinfo.cstime To get a feel for the numbers and the amount of information, use -f with pminfo in place of -dT. > Please tell me how i can get the process metrics programmatically using > pcp libraries.DO i have to write a PMDA or PMAPI?? I expect you will need to write a client program that uses the PMAPI to connect to PMCD and fetch the desired proc.* metrics for each process ... if you are interested in _recent_ CPU usage then you'll have to include periodic fetching or fetch-sleep-fetch because the CPU usage metrics are free-running counters that return CPU time used since the process began ... pmclient and pmstat are examples of programs that fetch a small number of metrics using the PMAPI, I'd suggest using pmclient as source reference. From owner-pcp@oss.sgi.com Fri Jul 6 07:48:29 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f66EmTm05097 for pcp-outgoing; Fri, 6 Jul 2001 07:48:29 -0700 Received: from rigel.cis.ksu.edu (root@rigel.cis.ksu.edu [129.130.10.65]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f66EmSV05094 for ; Fri, 6 Jul 2001 07:48:28 -0700 Received: from pollux.cis.ksu.edu (mst9696@pollux.cis.ksu.edu [129.130.10.33]) by rigel.cis.ksu.edu (8.9.1/8.9.1/000517) with ESMTP id JAA03031; Fri, 6 Jul 2001 09:48:25 -0500 (CDT) Received: from localhost (mst9696@localhost) by pollux.cis.ksu.edu (8.9.1/8.9.1/000517) with ESMTP id JAA24625; Fri, 6 Jul 2001 09:48:25 -0500 (CDT) X-Authentication-Warning: pollux.cis.ksu.edu: mst9696 owned process doing -bs Date: Fri, 6 Jul 2001 09:48:25 -0500 (CDT) From: Madhu Sudhan R Tera To: kenmcd@sgi.com cc: pcp@oss.sgi.com Subject: Re: getting the currently running process metrics In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pcp@oss.sgi.com Precedence: bulk Hello! Thank you for your quick response. As mentioned in my previous mail,i have recently installed PCP. But is there a procedure to pass process ids and get the metrics while writing a PMAPI??And what is this process.psinfo.stime and others in pminfo command(As far as i understand its a namespace.Correct me if i am wrong!) Thank you, With regards, MadhuSudhan Reddy Tera -------------------------------------------------- |Department of Computing and Information Sciences, | |Kansas State University, | |Manhattan, KS 66502. | |ph no:785-537-3276 | -------------------------------------------------- On Fri, 6 Jul 2001, Ken McDonell wrote: > On Thu, 5 Jul 2001, Madhu Sudhan R Tera wrote: > > > Hello all! > > I am new to this mailing list! > > Welcome. > > > Currently i am working on load balancing in beowulf cluster. > > > > The proposed procedure is to get the metrics for all the running > > processes from PCP and provide these metrics as input to the loadbalancing > > algorithm, which evaluates these metrics and decides which process to be > > migrated. > > Seems like a reasonable fit with PCP's capabilities. > > > But I dont find any facility to get the process metrics like CPU usage by > > each running process or memorey required by each process.pmem tool is not > > available in non-commercial versions of PCP,i guess! > > The reason pmem is not there is that compared to IRIX, Linux does not export the same > (nor as detailed) information about the address space of each process. > > But there are some stats ... check out > > pminfo -dT proc.memory > > Similarly the CPU usage stats are covered by > > pminfo -dT proc.psinfo.utime proc.psinfo.stime proc.psinfo.cutime \ > proc.psinfo.cstime > > To get a feel for the numbers and the amount of information, use -f > with pminfo in place of -dT. > > > Please tell me how i can get the process metrics programmatically using > > pcp libraries.DO i have to write a PMDA or PMAPI?? > > I expect you will need to write a client program that uses the PMAPI to > connect to PMCD and fetch the desired proc.* metrics for each process > ... if you are interested in _recent_ CPU usage then you'll have to > include periodic fetching or fetch-sleep-fetch because the CPU usage > metrics are free-running counters that return CPU time used since the > process began ... pmclient and pmstat are examples of programs that > fetch a small number of metrics using the PMAPI, I'd suggest using > pmclient as source reference. > > From owner-pcp@oss.sgi.com Tue Jul 10 08:32:19 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6AFWJF28386 for pcp-outgoing; Tue, 10 Jul 2001 08:32:19 -0700 Received: from mailgw2.netvision.net.il (mailgw2.netvision.net.il [194.90.1.9]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6AFWGV28382 for ; Tue, 10 Jul 2001 08:32:17 -0700 Received: from mail.exanet.co.il ([212.143.73.99]) by mailgw2.netvision.net.il (8.9.3/8.9.3) with ESMTP id SAA02709 for ; Tue, 10 Jul 2001 18:34:32 +0300 (IDT) Received: from cat (fw1.exanet.co.il [212.143.73.98]) by mail.exanet.co.il (8.11.0/8.11.0) with SMTP id f6AFScg23616 for ; Tue, 10 Jul 2001 18:28:38 +0300 Reply-To: From: "gilly" To: Subject: porting pmapi to windows Date: Tue, 10 Jul 2001 18:30:47 +0200 Message-ID: <01f701c1095d$ad2312d0$2a04000a@exanet> MIME-Version: 1.0 Content-Type: text/plain; charset="windows-1255" Content-Transfer-Encoding: 7bit X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook CWS, Build 9.0.2416 (9.0.2910.0) Importance: Normal X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400 Sender: owner-pcp@oss.sgi.com Precedence: bulk Can I use the pmapi for monitoring from a windows machine (the pmcd & pmdas set on linux machines only)? Is there any special porting that need to be done? thanks, gilly From owner-pcp@oss.sgi.com Tue Jul 10 18:48:42 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6B1mgr23884 for pcp-outgoing; Tue, 10 Jul 2001 18:48:42 -0700 Received: from clmboh1-smtp3.columbus.rr.com (clmboh1-smtp3.columbus.rr.com [65.24.0.112]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6B1mfV23877 for ; Tue, 10 Jul 2001 18:48:41 -0700 Received: from double (l1-1a089.neo.rr.com [24.93.251.89]) by clmboh1-smtp3.columbus.rr.com (8.11.2/8.11.2) with SMTP id f6B1jDZ17197 for ; Tue, 10 Jul 2001 21:45:14 -0400 (EDT) Date: Tue, 10 Jul 2001 21:48:31 -0400 From: Timothy Reaves To: PCP Subject: GUI? Message-Id: <20010710214831.08dff38a.treaves@silverfields.com> Organization: Silverfields X-Mailer: Sylpheed version 0.5.0 (GTK+ 1.2.8; i586-pc-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-pcp@oss.sgi.com Precedence: bulk Has any GUI replaced PCPMON? As it's dead, I'd think one would be useful. From owner-pcp@oss.sgi.com Tue Jul 10 18:55:33 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6B1tXS24490 for pcp-outgoing; Tue, 10 Jul 2001 18:55:33 -0700 Received: from imrelay.zambeel.com ([63.89.188.9]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6B1tWV24481 for ; Tue, 10 Jul 2001 18:55:32 -0700 Received: from xchange.zambeel.com ([63.89.188.10]) by imrelay.zambeel.com (8.11.0/8.11.0) with ESMTP id f6B1ubI02847; Tue, 10 Jul 2001 18:56:37 -0700 Received: by exchange.zambeel.com with Internet Mail Service (5.5.2653.19) id ; Tue, 10 Jul 2001 18:55:24 -0700 Message-ID: <2B8089144916D411896D00D0B73C8353015C3ADF@exchange.zambeel.com> From: Keith Hankin To: "'Timothy Reaves'" , PCP Subject: RE: GUI? Date: Tue, 10 Jul 2001 18:55:16 -0700 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2653.19) Content-Type: text/plain; charset="ISO-8859-1" Sender: owner-pcp@oss.sgi.com Precedence: bulk What do you mean "It's dead"? Is no one working on it anymore? > -----Original Message----- > From: Timothy Reaves [mailto:treaves@silverfields.com] > Sent: Tuesday, July 10, 2001 6:49 PM > To: PCP > Subject: GUI? > > > Has any GUI replaced PCPMON? As it's dead, I'd think > one would be useful. > From owner-pcp@oss.sgi.com Tue Jul 10 19:01:49 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6B21nP25021 for pcp-outgoing; Tue, 10 Jul 2001 19:01:49 -0700 Received: from clmboh1-smtp3.columbus.rr.com (clmboh1-smtp3.columbus.rr.com [65.24.0.112]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6B21lV25017 for ; Tue, 10 Jul 2001 19:01:48 -0700 Received: from double (l1-1a089.neo.rr.com [24.93.251.89]) by clmboh1-smtp3.columbus.rr.com (8.11.2/8.11.2) with SMTP id f6B1wFZ26200 for ; Tue, 10 Jul 2001 21:58:20 -0400 (EDT) Date: Tue, 10 Jul 2001 22:01:32 -0400 From: Timothy Reaves To: pcp@oss.sgi.com Subject: Re: GUI? Message-Id: <20010710220132.561b8da0.treaves@silverfields.com> In-Reply-To: <2B8089144916D411896D00D0B73C8353015C3ADF@exchange.zambeel.com> References: <2B8089144916D411896D00D0B73C8353015C3ADF@exchange.zambeel.com> Organization: Silverfields X-Mailer: Sylpheed version 0.5.0 (GTK+ 1.2.8; i586-pc-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-pcp@oss.sgi.com Precedence: bulk Unfortunatly. The last release was about eight months ago, and getting it to build on a current (SuSE 7.1) system is just about impossible. Several of the libraries it uses (XML & GIF) are out-of-date, and downgrading would break other things. On Tue, 10 Jul 2001 18:55:16 -0700 Keith Hankin wrote: > What do you mean "It's dead"? Is no one working on it anymore? > > > > > Has any GUI replaced PCPMON? As it's dead, I'd think > > one would be useful. > > From owner-pcp@oss.sgi.com Tue Jul 10 22:27:24 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6B5RO611624 for pcp-outgoing; Tue, 10 Jul 2001 22:27:24 -0700 Received: from sgi.com (sgi.SGI.COM [192.48.153.1]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6B5RNV11617 for ; Tue, 10 Jul 2001 22:27:23 -0700 Received: from larry.melbourne.sgi.com (larry.melbourne.sgi.com [134.14.52.130]) by sgi.com (980327.SGI.8.8.8-aspam/980304.SGI-aspam: SGI does not authorize the use of its proprietary systems or networks for unsolicited or bulk email from the Internet.) via SMTP id WAA07261 for ; Tue, 10 Jul 2001 22:27:22 -0700 (PDT) mail_from (kenmcd@melbourne.sgi.com) From: kenmcd@melbourne.sgi.com Received: from [192.82.201.208] ([192.82.201.208]) by larry.melbourne.sgi.com (950413.SGI.8.6.12/950213.SGI.AUTOCF) via ESMTP id PAA18659; Wed, 11 Jul 2001 15:26:00 +1000 Date: Wed, 11 Jul 2001 15:25:36 +1000 (EST) X-Sender: kenmcd@masala.engr.sgi.com Reply-To: kenmcd@melbourne.sgi.com To: lemming@arthur.plbohnice.cz cc: pcp@oss.sgi.com Subject: Re: GUI? (fwd) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pcp@oss.sgi.com Precedence: bulk Since pcpmon was built by someone outside SGI ... Michal are you still lurking on the list? ---------- Forwarded message ---------- Date: Tue, 10 Jul 2001 22:01:32 -0400 From: Timothy Reaves To: pcp@oss.sgi.com Subject: Re: GUI? Unfortunatly. The last release was about eight months ago, and getting it to build on a current (SuSE 7.1) system is just about impossible. Several of the libraries it uses (XML & GIF) are out-of-date, and downgrading would break other things. On Tue, 10 Jul 2001 18:55:16 -0700 Keith Hankin wrote: > What do you mean "It's dead"? Is no one working on it anymore? > > > > > Has any GUI replaced PCPMON? As it's dead, I'd think > > one would be useful. > > From owner-pcp@oss.sgi.com Wed Jul 11 00:25:01 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6B7P1W22887 for pcp-outgoing; Wed, 11 Jul 2001 00:25:01 -0700 Received: from vulcan.alchemic.com (cb842350-d.rchstr1.mn.home.com [24.178.208.188]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6B7OxV22882 for ; Wed, 11 Jul 2001 00:24:59 -0700 Received: from uberh4x0r.org (terra.alchemic.com [192.168.0.9]) by vulcan.alchemic.com (8.11.0/8.11.0) with ESMTP id f6B7OkG23077; Wed, 11 Jul 2001 02:24:51 -0500 Message-ID: <3B4BFF3E.80309@uberh4x0r.org> Date: Wed, 11 Jul 2001 02:24:46 -0500 From: Dean Johnson User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:0.9.2) Gecko/20010628 X-Accept-Language: en-us MIME-Version: 1.0 To: Timothy Reaves CC: pcp@oss.sgi.com Subject: Re: GUI? References: <2B8089144916D411896D00D0B73C8353015C3ADF@exchange.zambeel.com> <20010710220132.561b8da0.treaves@silverfields.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-pcp@oss.sgi.com Precedence: bulk Timothy Reaves wrote: > Unfortunatly. The last release was about eight months ago, and getting it to build on a current (SuSE 7.1) system is just about impossible. Several of the libraries it uses (XML & GIF) are out-of-date, and downgrading would break other things. > > I just built it on a minimally mangled Redhat 7.1 (SGI's xfs version) with Ximian. No problems building it and only a couple of warnings. I don't have a Suse 7.1 system, only a 6.4 system. -Dean From owner-pcp@oss.sgi.com Wed Jul 11 00:33:59 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6B7Xx223664 for pcp-outgoing; Wed, 11 Jul 2001 00:33:59 -0700 Received: from arthur.plbohnice.cz ([194.108.220.193]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6B7XuV23657 for ; Wed, 11 Jul 2001 00:33:57 -0700 Received: (from lemming@localhost) by arthur.plbohnice.cz (8.9.3/8.10.1) id JAA21595 for pcp@oss.sgi.com; Wed, 11 Jul 2001 09:33:56 +0200 Date: Wed, 11 Jul 2001 09:33:55 +0200 From: Michal Kara To: pcp@oss.sgi.com Subject: Re: GUI? Message-ID: <20010711093355.A21441@arthur.plbohnice.cz> References: <2B8089144916D411896D00D0B73C8353015C3ADF@exchange.zambeel.com> <20010710220132.561b8da0.treaves@silverfields.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 1.0.1i In-Reply-To: <20010710220132.561b8da0.treaves@silverfields.com>; from treaves@silverfields.com on Tue, Jul 10, 2001 at 10:01:32PM -0400 Sender: owner-pcp@oss.sgi.com Precedence: bulk > Unfortunatly. The last release was about eight months ago, and getting > it to build on a current (SuSE 7.1) system is just about impossible. > Several of the libraries it uses (XML & GIF) are out-of-date, and > downgrading would break other things. Sorry. I have been busy with other things and I did not update distribution on the web page. A new version (1.3.2) is now available from http://k332.feld.cvut.cz/~lemming/projects/pcpmon.html Snap from the changelog is below. Logarhitmis axes still do not work properly - that's why I didn't release the version before. This version should be able to automagically work with any version of libgd/libxml. Michal 11-Jul-2001 1.3.2 lemming - Logarithmic axes added but currently disabled because they do not work properly - Oversampling added - Number of characters in labels enlarged - When in shot mode, error messages are drawn to the image. 23-Nov-2000 1.3.1 lemming - Program now uses gdImageGif when gdImagePng is not available - Works with both libxml1 and libxml2 (version is autodetected) 3-Nov-2000 1.3.0 lemming - Archive mode added - Screenshot ability added - Command line mode added (to generate just screen shot) - Program rewritten for libxml2 15-May-2000 1.2.0 lemming From owner-pcp@oss.sgi.com Wed Jul 11 05:08:12 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6BC8CI17245 for pcp-outgoing; Wed, 11 Jul 2001 05:08:12 -0700 Received: from yog-sothoth.sgi.com (eugate.sgi.com [192.48.160.10]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6BC88V17218 for ; Wed, 11 Jul 2001 05:08:08 -0700 Received: from larry.melbourne.sgi.com (larry.melbourne.sgi.com [134.14.52.130]) by yog-sothoth.sgi.com (980305.SGI.8.8.8-aspam-6.2/980304.SGI-aspam-europe) via SMTP id OAA19139 for ; Wed, 11 Jul 2001 14:08:05 +0200 (CEST) mail_from (kenmcd@melbourne.sgi.com) From: kenmcd@melbourne.sgi.com Received: from [192.82.201.77] ([192.82.201.77]) by larry.melbourne.sgi.com (950413.SGI.8.6.12/950213.SGI.AUTOCF) via ESMTP id WAA20662; Wed, 11 Jul 2001 22:06:43 +1000 Date: Wed, 11 Jul 2001 22:06:35 +1000 (EST) X-Sender: kenmcd@masala.engr.sgi.com Reply-To: kenmcd@melbourne.sgi.com To: Madhu Sudhan R Tera cc: pcp@oss.sgi.com Subject: Re: getting the currently running process metrics In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pcp@oss.sgi.com Precedence: bulk On Fri, 6 Jul 2001, Madhu Sudhan R Tera wrote: > Hello! > Thank you for your quick response. > > As mentioned in my previous mail,i have recently installed PCP. > > But is there a procedure to pass process ids and get the metrics > while writing a PMAPI?? You'll need pmLookupDesc(3) to get the instance domain number associated with the metrics you are interested in, and then pmGetInDom(3) and/or pmLookupInDom(3) to process the instance domain that encodes the process ids. Then pmAddProfile(3) and pmDelProfile(3) to build/update a "profile". And finally pmFetch(3) to actually do the work. You can see an example of all of this in the source for pmval, e.g. the command $ pmval -i 920 proc.psinfo.utime reports the user CPU time for process id 920. To get some idea of the interaction between the client (pmval in this case) and pmcd below the PMAPI, use the PDU tracing option: $ pmval -i 920 -D pdu proc.psinfo.utime > ... And what is this process.psinfo.stime and > others in pminfo command(As far as i understand its a namespace.Correct me > if i am wrong!) I'm not going to explain them, that's the PMDA writer's job ... the -T option to pminfo requests the help text for the named metrics, e.g. $ pminfo -T proc.psinfo.utime proc.psinfo.utime Help: time (in ms) spent executing user code since process started From owner-pcp@oss.sgi.com Wed Jul 11 06:04:06 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6BD46O25090 for pcp-outgoing; Wed, 11 Jul 2001 06:04:06 -0700 Received: from clmboh1-smtp3.columbus.rr.com (clmboh1-smtp3.columbus.rr.com [65.24.0.112]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6BD44V25080 for ; Wed, 11 Jul 2001 06:04:05 -0700 Received: from double (l1-1a089.neo.rr.com [24.93.251.89]) by clmboh1-smtp3.columbus.rr.com (8.11.2/8.11.2) with SMTP id f6BD0bZ10879 for ; Wed, 11 Jul 2001 09:00:38 -0400 (EDT) Date: Wed, 11 Jul 2001 09:03:51 -0400 From: Timothy Reaves To: pcp@oss.sgi.com Subject: Re: GUI? Message-Id: <20010711090351.5b9401f9.treaves@silverfields.com> In-Reply-To: <20010711093355.A21441@arthur.plbohnice.cz> References: <2B8089144916D411896D00D0B73C8353015C3ADF@exchange.zambeel.com> <20010710220132.561b8da0.treaves@silverfields.com> <20010711093355.A21441@arthur.plbohnice.cz> Organization: Silverfields X-Mailer: Sylpheed version 0.5.0 (GTK+ 1.2.8; i586-pc-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-pcp@oss.sgi.com Precedence: bulk On Wed, 11 Jul 2001 09:33:55 +0200 Michal Kara wrote: > Sorry. I have been busy with other things and I did not update distribution on > the web page. > > A new version (1.3.2) is now available from > http://k332.feld.cvut.cz/~lemming/projects/pcpmon.html > > Snap from the changelog is below. Logarhitmis axes still do not work properly > - that's why I didn't release the version before. This version should be able to > automagically work with any version of libgd/libxml. > Other than some warnings, it does. Thank you! From owner-pcp@oss.sgi.com Wed Jul 11 06:04:42 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6BD4gd25149 for pcp-outgoing; Wed, 11 Jul 2001 06:04:42 -0700 Received: from sgi.com (sgi.SGI.COM [192.48.153.1]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6BD4fV25146 for ; Wed, 11 Jul 2001 06:04:41 -0700 Received: from larry.melbourne.sgi.com (larry.melbourne.sgi.com [134.14.52.130]) by sgi.com (980327.SGI.8.8.8-aspam/980304.SGI-aspam: SGI does not authorize the use of its proprietary systems or networks for unsolicited or bulk email from the Internet.) via SMTP id GAA08344 for ; Wed, 11 Jul 2001 06:04:30 -0700 (PDT) mail_from (kenmcd@melbourne.sgi.com) From: kenmcd@melbourne.sgi.com Received: from [192.82.201.77] ([192.82.201.77]) by larry.melbourne.sgi.com (950413.SGI.8.6.12/950213.SGI.AUTOCF) via ESMTP id XAA20960; Wed, 11 Jul 2001 23:03:07 +1000 Date: Wed, 11 Jul 2001 23:02:59 +1000 (EST) X-Sender: kenmcd@masala.engr.sgi.com Reply-To: kenmcd@melbourne.sgi.com To: gilly cc: pcp@oss.sgi.com Subject: Re: porting pmapi to windows In-Reply-To: <01f701c1095d$ad2312d0$2a04000a@exanet> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pcp@oss.sgi.com Precedence: bulk On Tue, 10 Jul 2001, gilly wrote: > Can I use the pmapi for monitoring from a windows machine (the pmcd & pmdas > set on linux machines only)? > Is there any special porting that need to be done? There is lots of porting (we've done some of this in the past, but did not complete a fully-functional PCP for anything-from-Redmond). The main hurdles are: - you'll need a real C compiler and run-time environment that makes your Windows system look like an ANSI C environment - you'll have to develop the Windows PMDA from scratch From owner-pcp@oss.sgi.com Fri Jul 13 19:00:16 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6E20Gx11590 for pcp-outgoing; Fri, 13 Jul 2001 19:00:16 -0700 Received: from ex1.ncsa.uiuc.edu (ex1.ncsa.uiuc.edu [141.142.2.9]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6E20FV11587 for ; Fri, 13 Jul 2001 19:00:15 -0700 Received: from mx1.ncsa.uiuc.edu (mx1.ncsa.uiuc.edu [141.142.2.8]) by ex1.ncsa.uiuc.edu (8.11.0/8.11.0) with ESMTP id f6E20EL18699 for ; Fri, 13 Jul 2001 21:00:14 -0500 (CDT) X-Envelope-From: abailey@ncsa.uiuc.edu X-Envelope-To: Received: from osage.ncsa.uiuc.edu (osage.ncsa.uiuc.edu [141.142.2.56]) by mx1.ncsa.uiuc.edu (8.11.0/8.11.0) with ESMTP id f6E20CP15129 for ; Fri, 13 Jul 2001 21:00:13 -0500 (CDT) Received: from localhost (abailey@localhost) by osage.ncsa.uiuc.edu (8.9.3/8.9.3) with ESMTP id VAA02297 for ; Fri, 13 Jul 2001 21:00:12 -0500 Date: Fri, 13 Jul 2001 21:00:11 -0500 (CDT) From: Alan Bailey To: Subject: why not install as DSO? Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pcp@oss.sgi.com Precedence: bulk I've got a simple question. Why is the default to install pmdas as daemons instead of as a DSO? I quote from pmdaproc.sh: # Install control variables # Can install as DSO? dso_opt=false I kinda like installing them as DSO's because it puts everything into one process. It might also end up taking up less memory. Which one is recommended, and why? Or is this something that is figured out at install time, and since I'm on a linux system, it thinks it shouldn't install it as a DSO? Just wondering. Alan -- Alan Bailey From owner-pcp@oss.sgi.com Fri Jul 13 21:26:04 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6E4Q4r19271 for pcp-outgoing; Fri, 13 Jul 2001 21:26:04 -0700 Received: from sgi.com (sgi.SGI.COM [192.48.153.1]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6E4Q2V19267 for ; Fri, 13 Jul 2001 21:26:02 -0700 Received: from larry.melbourne.sgi.com (larry.melbourne.sgi.com [134.14.52.130]) by sgi.com (980327.SGI.8.8.8-aspam/980304.SGI-aspam: SGI does not authorize the use of its proprietary systems or networks for unsolicited or bulk email from the Internet.) via SMTP id VAA07580 for ; Fri, 13 Jul 2001 21:25:59 -0700 (PDT) mail_from (nathans@wobbly.melbourne.sgi.com) Received: from wobbly.melbourne.sgi.com (wobbly.melbourne.sgi.com [134.14.55.135]) by larry.melbourne.sgi.com (950413.SGI.8.6.12/950213.SGI.AUTOCF) via ESMTP id OAA10632; Sat, 14 Jul 2001 14:24:44 +1000 Received: (from nathans@localhost) by wobbly.melbourne.sgi.com (SGI-8.9.3/8.9.3) id OAA11519; Sat, 14 Jul 2001 14:24:43 +1000 (EST) From: "Nathan Scott" Message-Id: <10107141424.ZM211498@wobbly.melbourne.sgi.com> Date: Sat, 14 Jul 2001 14:24:41 -0500 In-Reply-To: Alan Bailey "why not install as DSO?" (Jul 13, 9:00pm) References: X-Mailer: Z-Mail (3.2.3 08feb96 MediaMail) To: Alan Bailey , Subject: Re: why not install as DSO? Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-pcp@oss.sgi.com Precedence: bulk hi Alan, On Jul 13, 9:00pm, Alan Bailey wrote: > Subject: why not install as DSO? > I've got a simple question. Why is the default to install pmdas as > daemons instead of as a DSO? I quote from pmdaproc.sh: > > # Install control variables > # Can install as DSO? > dso_opt=false > > I kinda like installing them as DSO's because it puts everything into one > process. It might also end up taking up less memory. Which one is > recommended, and why? Or is this something that is figured out at install > time, and since I'm on a linux system, it thinks it shouldn't install it > as a DSO? > > Just wondering. > One thing to bear in mind is that pmcd is single threaded by design, so a PMDA which takes some time to fetch new data can cause pmcd to respond more slowly - eg. if the fetch mechanism has to block while obtaining data. The kernel PMDAs (for both IRIX + Linux) tend to use system calls to fetch their data and are very lightweight - other PMDAs are not quite so cheap. Another factor is that a separate process is self-contained, so a crash in a non-DSO PMDA won't cause pmcd to die & pmcd can continue on gracefully serving up other metrics in this situation. There is little difference between the Install scripts on Linux and IRIX, so the choice of DSO/daemon usually defaults to daemon (on all platforms) to try to improve parallelism and robustness. cheers. -- Nathan From owner-pcp@oss.sgi.com Fri Jul 13 22:00:39 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6E50dA20652 for pcp-outgoing; Fri, 13 Jul 2001 22:00:39 -0700 Received: from tomts8-srv.bellnexxia.net (tomts8.bellnexxia.net [209.226.175.52]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6E50cV20649 for ; Fri, 13 Jul 2001 22:00:38 -0700 Received: from planux.com ([65.92.128.127]) by tomts8-srv.bellnexxia.net (InterMail vM.4.01.03.16 201-229-121-116-20010115) with ESMTP id <20010714050031.BKEJ3306.tomts8-srv.bellnexxia.net@planux.com>; Sat, 14 Jul 2001 01:00:31 -0400 Message-ID: <3B4FD210.80506@planux.com> Date: Sat, 14 Jul 2001 01:01:04 -0400 From: Anas Nashif User-Agent: Mozilla/5.0 (X11; U; Linux 2.2.19 i686; en-US; rv:0.9) Gecko/20010511 X-Accept-Language: en MIME-Version: 1.0 To: Timothy Reaves CC: pcp@oss.sgi.com Subject: Re: GUI? References: <2B8089144916D411896D00D0B73C8353015C3ADF@exchange.zambeel.com> <20010710220132.561b8da0.treaves@silverfields.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-pcp@oss.sgi.com Precedence: bulk Timothy Reaves wrote: > Unfortunatly. The last release was about eight months ago, and getting it to build on a current (SuSE 7.1) system is just about impossible. Several of the libraries it uses (XML & GIF) are out-of-date, and downgrading would break other things. > > Why do you need to build it with SuSE? It is available on the distribution already. Anas > On Tue, 10 Jul 2001 18:55:16 -0700 > Keith Hankin wrote: > > >>What do you mean "It's dead"? Is no one working on it anymore? >> >> >>> Has any GUI replaced PCPMON? As it's dead, I'd think >>>one would be useful. >>> >>> From owner-pcp@oss.sgi.com Mon Jul 16 02:42:40 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6G9geu15753 for pcp-outgoing; Mon, 16 Jul 2001 02:42:40 -0700 Received: from mailgw1.netvision.net.il (mailgw.netvision.net.il [194.90.1.14]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6G9gbV15746 for ; Mon, 16 Jul 2001 02:42:37 -0700 Received: from mail.exanet.co.il ([212.143.73.99]) by mailgw1.netvision.net.il (8.9.3/8.9.3) with ESMTP id MAA10349 for ; Mon, 16 Jul 2001 12:42:33 +0300 (IDT) Received: from cat (fw1.exanet.co.il [212.143.73.98]) by mail.exanet.co.il (8.11.0/8.11.0) with SMTP id f6G9c9g32497 for ; Mon, 16 Jul 2001 12:38:09 +0300 Reply-To: From: "gilly" To: Subject: a PMDA using another PMDA Date: Mon, 16 Jul 2001 12:40:51 +0200 Message-ID: <003b01c10de3$c9218900$2a04000a@exanet> MIME-Version: 1.0 Content-Type: text/plain; charset="windows-1255" Content-Transfer-Encoding: 7bit X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook CWS, Build 9.0.2416 (9.0.2910.0) X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400 Importance: Normal Sender: owner-pcp@oss.sgi.com Precedence: bulk Is it possible to write a PMDA which its metrics are accumulative metrics of another PMDA from other hosts? for example - a PMDA that calculates all RPC metrics of 5 hosts it monitors, therefore inside the 'total-RPC' PMDA there's a call to pmFetch for the rpc metrics (from linux PMDA) from the other hosts (the context is - PM_CONTEXT_HOST). Actually the call to 'pmFetch' for the accumulative metrics calls pmFetch for each host - for the calculated metrics. Does it matter if the PMDA's are installed as DSO? thanks in advance, gilly From owner-pcp@oss.sgi.com Mon Jul 16 19:15:59 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6H2Fxl06289 for pcp-outgoing; Mon, 16 Jul 2001 19:15:59 -0700 Received: from sgi.com (sgi.SGI.COM [192.48.153.1]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6H2FvV06286 for ; Mon, 16 Jul 2001 19:15:57 -0700 Received: from larry.melbourne.sgi.com (larry.melbourne.sgi.com [134.14.52.130]) by sgi.com (980327.SGI.8.8.8-aspam/980304.SGI-aspam: SGI does not authorize the use of its proprietary systems or networks for unsolicited or bulk email from the Internet.) via SMTP id TAB08772 for ; Mon, 16 Jul 2001 19:14:03 -0700 (PDT) mail_from (nathans@wobbly.melbourne.sgi.com) Received: from wobbly.melbourne.sgi.com (wobbly.melbourne.sgi.com [134.14.55.135]) by larry.melbourne.sgi.com (950413.SGI.8.6.12/950213.SGI.AUTOCF) via ESMTP id MAA29023; Tue, 17 Jul 2001 12:12:41 +1000 Received: (from nathans@localhost) by wobbly.melbourne.sgi.com (SGI-8.9.3/8.9.3) id MAA02982; Tue, 17 Jul 2001 12:12:38 +1000 (AEST) From: "Nathan Scott" Message-Id: <10107171212.ZM204950@wobbly.melbourne.sgi.com> Date: Tue, 17 Jul 2001 12:12:36 +1000 In-Reply-To: "gilly" "a PMDA using another PMDA" (Jul 16, 12:40pm) References: <003b01c10de3$c9218900$2a04000a@exanet> X-Mailer: Z-Mail (3.2.3 08feb96 MediaMail) To: , Subject: Re: a PMDA using another PMDA Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-pcp@oss.sgi.com Precedence: bulk hi, On Jul 16, 12:40pm, gilly wrote: > Subject: a PMDA using another PMDA > Is it possible to write a PMDA which its metrics are accumulative metrics of > another PMDA from other hosts? for example - a PMDA that calculates all RPC > metrics of 5 hosts it monitors, therefore inside the 'total-RPC' PMDA > there's a call to pmFetch for the rpc metrics (from linux PMDA) from the > other hosts (the context is - PM_CONTEXT_HOST). Actually the call to > 'pmFetch' for the accumulative metrics calls pmFetch for each host - for the > calculated metrics. Yes, that is possible. In fact on IRIX & in pcp-pro there is a configurable "summary" PMDA which used pmie(1) to do just this sort of thing. > Does it matter if the PMDA's are installed as DSO? Do you mean in general or in this particular instance? See my mail to the list from the other day for some thoughts on this topic. cheers. -- Nathan From owner-pcp@oss.sgi.com Mon Jul 16 20:20:04 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6H3K4Y07353 for pcp-outgoing; Mon, 16 Jul 2001 20:20:04 -0700 Received: from deliverator.sgi.com (deliverator.sgi.com [204.94.214.10]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6H3K3V07349 for ; Mon, 16 Jul 2001 20:20:03 -0700 Received: from larry.melbourne.sgi.com (larry.melbourne.sgi.com [134.14.52.130]) by deliverator.sgi.com (980309.SGI.8.8.8-aspam-6.2/980310.SGI-aspam) via SMTP id UAA15353 for ; Mon, 16 Jul 2001 20:18:43 -0700 (PDT) mail_from (markgw@sgi.com) Received: from sherman.melbourne.sgi.com (sherman.melbourne.sgi.com [134.14.55.175]) by larry.melbourne.sgi.com (950413.SGI.8.6.12/950213.SGI.AUTOCF) via ESMTP id NAA29411; Tue, 17 Jul 2001 13:17:31 +1000 Date: Tue, 17 Jul 2001 13:17:31 +1000 (EST) From: Mark Goodwin X-Sender: markgw@sherman.melbourne.sgi.com To: Nathan Scott cc: gilly@exanet.com, pcp@oss.sgi.com Subject: Re: a PMDA using another PMDA In-Reply-To: <10107171212.ZM204950@wobbly.melbourne.sgi.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pcp@oss.sgi.com Precedence: bulk On Tue, 17 Jul 2001, Nathan Scott wrote: > Yes, that is possible. In fact on IRIX & in pcp-pro there > is a configurable "summary" PMDA which used pmie(1) to do > just this sort of thing. Specifically, see the -x flag to pmie. This tells pmie to periodically send PDUs rather than print results. A really nice feature of pmie is that the PDU's sent by pmie can contain metric values that have been derived as arbitrary expressions of other metric values from one or multiple hosts, even localhost. Your PCP agent should fork, then dup stdin/stdout and exec pmie -x. Your agent also needs to add pmie's file descriptor to it's select mask. To do this you need to provide your own pmdaMainLoop; read the man page for pmdaMain(3) and then base your agent's main loop on the code in src/libpcp_pmda/src/mainloop.c. When pmie sends something, your agent will wake up, read the new value(s) and cache locally. > > > Does it matter if the PMDA's are installed as DSO? > > Do you mean in general or in this particular instance? See > my mail to the list from the other day for some thoughts on > this topic. I would not advise a DSO agent that fiddles with the select mask. (pmcd's select loop is already complicated enough). -- Mark From owner-pcp@oss.sgi.com Mon Jul 16 20:54:15 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6H3sFt07816 for pcp-outgoing; Mon, 16 Jul 2001 20:54:15 -0700 Received: from pneumatic-tube.sgi.com (pneumatic-tube.sgi.com [204.94.214.22]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6H3sEV07813 for ; Mon, 16 Jul 2001 20:54:14 -0700 Received: from larry.melbourne.sgi.com (larry.melbourne.sgi.com [134.14.52.130]) by pneumatic-tube.sgi.com (980327.SGI.8.8.8-aspam/980310.SGI-aspam) via SMTP id UAA03625 for ; Mon, 16 Jul 2001 20:51:45 -0700 (PDT) mail_from (markgw@sgi.com) Received: from sherman.melbourne.sgi.com (sherman.melbourne.sgi.com [134.14.55.175]) by larry.melbourne.sgi.com (950413.SGI.8.6.12/950213.SGI.AUTOCF) via ESMTP id NAA29603; Tue, 17 Jul 2001 13:52:54 +1000 Date: Tue, 17 Jul 2001 13:52:54 +1000 (EST) From: Mark Goodwin X-Sender: markgw@sherman.melbourne.sgi.com To: kenmcd@melbourne.sgi.com cc: gilly , pcp@oss.sgi.com Subject: Re: porting pmapi to windows In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pcp@oss.sgi.com Precedence: bulk On Wed, 11 Jul 2001 kenmcd@melbourne.sgi.com wrote: > On Tue, 10 Jul 2001, gilly wrote: > > > Can I use the pmapi for monitoring from a windows machine (the pmcd & pmdas > > set on linux machines only)? > > Is there any special porting that need to be done? > > There is lots of porting (we've done some of this in the past, but > did not complete a fully-functional PCP for anything-from-Redmond). > > The main hurdles are: > > - you'll need a real C compiler and run-time environment that makes > your Windows system look like an ANSI C environment The cygwin environment from Redhat/Cygnus is probably the most suitable. This environment supports autoconf/configure and the GNU toolchain on windows, and provides a pretty much complete Unix libc and system call API over win32. Get it from http://sources.redhat.com/cygwin/ > > - you'll have to develop the Windows PMDA from scratch > There is a win32 API for extracting perf stats out of the registry. I don't know if cygwin apps can call it or not. Another issue is that win32 does not have a single/global file system root. From memory, cygwin has a workaround for this. If possible, please share anything you come up with! Good luck! -- Mark From owner-pcp@oss.sgi.com Wed Jul 18 00:33:13 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6I7XDh08984 for pcp-outgoing; Wed, 18 Jul 2001 00:33:13 -0700 Received: from pneumatic-tube.sgi.com (pneumatic-tube.sgi.com [204.94.214.22]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6I7XBV08981 for ; Wed, 18 Jul 2001 00:33:11 -0700 Received: from larry.melbourne.sgi.com (larry.melbourne.sgi.com [134.14.52.130]) by pneumatic-tube.sgi.com (980327.SGI.8.8.8-aspam/980310.SGI-aspam) via SMTP id AAA01902 for ; Wed, 18 Jul 2001 00:30:41 -0700 (PDT) mail_from (kenmcd@melbourne.sgi.com) From: kenmcd@melbourne.sgi.com Received: from [192.82.201.219] ([192.82.201.219]) by larry.melbourne.sgi.com (950413.SGI.8.6.12/950213.SGI.AUTOCF) via ESMTP id RAA09853; Wed, 18 Jul 2001 17:30:20 +1000 Date: Wed, 18 Jul 2001 17:30:29 +1000 (EST) X-Sender: kenmcd@fsgi627.americas.sgi.com Reply-To: kenmcd@melbourne.sgi.com To: Nathan Scott cc: Alan Bailey , pcp@oss.sgi.com Subject: Re: why not install as DSO? In-Reply-To: <10107141424.ZM211498@wobbly.melbourne.sgi.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pcp@oss.sgi.com Precedence: bulk Nathan addressed most of the issues. The one that I would add is that if you have a PMDA where you want to fetch _just_ the metrics from that PMDA (possibly with some IRIX/Linux metrics) _and_ you want a very fast sampling rate, then the DSO method is the one of choice. These pre-conditions are rare, so the default is to _not_ install as a DSO. Note that if you use the source of the existing PMDAs that link with libpcp_pmda as the base for your new plug in, then the code will compile to produce _both_ a DSO and a daemon PMDA ... this is not an accident ... the choice relates to a trade-off of run-time performance vs pmcd robustness vs PMDA flexibility (there are things like sproc/pthread that you'd have to be silly to use in a DSO PMDA). Hope this helps. From owner-pcp@oss.sgi.com Wed Jul 18 00:33:59 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6I7Xxm09037 for pcp-outgoing; Wed, 18 Jul 2001 00:33:59 -0700 Received: from yog-sothoth.sgi.com (eugate.sgi.com [192.48.160.10]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6I7XvV09033 for ; Wed, 18 Jul 2001 00:33:57 -0700 Received: from larry.melbourne.sgi.com (larry.melbourne.sgi.com [134.14.52.130]) by yog-sothoth.sgi.com (980305.SGI.8.8.8-aspam-6.2/980304.SGI-aspam-europe) via SMTP id JAA414913 for ; Wed, 18 Jul 2001 09:33:41 +0200 (CEST) mail_from (kenmcd@melbourne.sgi.com) From: kenmcd@melbourne.sgi.com Received: from [192.82.201.219] ([192.82.201.219]) by larry.melbourne.sgi.com (950413.SGI.8.6.12/950213.SGI.AUTOCF) via ESMTP id RAA09875; Wed, 18 Jul 2001 17:32:23 +1000 Date: Wed, 18 Jul 2001 17:32:32 +1000 (EST) X-Sender: kenmcd@fsgi627.americas.sgi.com Reply-To: kenmcd@melbourne.sgi.com To: gilly cc: pcp@oss.sgi.com Subject: Re: a PMDA using another PMDA In-Reply-To: <003b01c10de3$c9218900$2a04000a@exanet> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pcp@oss.sgi.com Precedence: bulk Nathan and Mark pointed to the summary PMDA as an existance proof. Just remember the PCP protocols are single-threaded, synchronous and with timeouts ... so be very careful about starvation and deadlock, e.g. consider the sequence: client -> pmcd -> summary PMDA -> pmie -> pmcd The summary PMDA needs two threads of control: 1. launch pmie and block waiting for new values to be returned from pmie 2. block on requests from pmcd, and return the most recently observed values from 1. If the nested PMDA is not calling pmcd, then this problem goes away, but you may have a mutex problem with the nested PMDA having to concurrently field requests from pmcd and the outermost PMDA ... remeber that pmcd runs DSO PMDA requests serially, but runs daemon PMDA requests in parallel. On Mon, 16 Jul 2001, gilly wrote: > Is it possible to write a PMDA which its metrics are accumulative metrics of > another PMDA from other hosts? for example - a PMDA that calculates all RPC > metrics of 5 hosts it monitors, therefore inside the 'total-RPC' PMDA > there's a call to pmFetch for the rpc metrics (from linux PMDA) from the > other hosts (the context is - PM_CONTEXT_HOST). Actually the call to > 'pmFetch' for the accumulative metrics calls pmFetch for each host - for the > calculated metrics. > Does it matter if the PMDA's are installed as DSO? > thanks in advance, gilly > From owner-pcp@oss.sgi.com Wed Jul 18 00:51:35 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6I7pZY10028 for pcp-outgoing; Wed, 18 Jul 2001 00:51:35 -0700 Received: from mailgw1.netvision.net.il (mailgw.netvision.net.il [194.90.1.14]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6I7pVV10023 for ; Wed, 18 Jul 2001 00:51:32 -0700 Received: from mail.exanet.co.il ([212.143.73.99]) by mailgw1.netvision.net.il (8.9.3/8.9.3) with ESMTP id KAA04596 for ; Wed, 18 Jul 2001 10:51:27 +0300 (IDT) Received: from cat (fw1.exanet.co.il [212.143.73.98]) by mail.exanet.co.il (8.11.0/8.11.0) with SMTP id f6I7kng25089 for ; Wed, 18 Jul 2001 10:46:49 +0300 Reply-To: From: "gilly" To: Subject: RE: a PMDA using another PMDA Date: Wed, 18 Jul 2001 10:49:49 +0200 Message-ID: <006e01c10f66$9b0e21c0$2a04000a@exanet> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook CWS, Build 9.0.2416 (9.0.2910.0) X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400 Importance: Normal In-Reply-To: Sender: owner-pcp@oss.sgi.com Precedence: bulk All the answers mentioned pmie, but what I'm trying to implement (and I'm still not sure I can do it) is to have a PMDA that performs sequential pmFetch calls - each for a different host, and sums the results as 'total-results'. Trying to run pminfo on this 'total' PMDA I get the following error: the metrics' values ARE calculated correctly, but when trying to get the instances' names (using pmGetInDom() in lookup()) - I get an "Unknown or illegal instance identifier" error. It seems that the second call (pmGetInDom, after pmFetch) to the PMDA on the other host doesn't reach its destination. Maybe I have a context problem? (I'm not sure I understand the contexts issue well.) Thanks, gilly -----Original Message----- From: kenmcd@melbourne.sgi.com [mailto:kenmcd@melbourne.sgi.com] Sent: Wednesday, July 18, 2001 9:33 AM To: gilly Cc: pcp@oss.sgi.com Subject: Re: a PMDA using another PMDA Nathan and Mark pointed to the summary PMDA as an existance proof. Just remember the PCP protocols are single-threaded, synchronous and with timeouts ... so be very careful about starvation and deadlock, e.g. consider the sequence: client -> pmcd -> summary PMDA -> pmie -> pmcd The summary PMDA needs two threads of control: 1. launch pmie and block waiting for new values to be returned from pmie 2. block on requests from pmcd, and return the most recently observed values from 1. If the nested PMDA is not calling pmcd, then this problem goes away, but you may have a mutex problem with the nested PMDA having to concurrently field requests from pmcd and the outermost PMDA ... remeber that pmcd runs DSO PMDA requests serially, but runs daemon PMDA requests in parallel. On Mon, 16 Jul 2001, gilly wrote: > Is it possible to write a PMDA which its metrics are accumulative metrics of > another PMDA from other hosts? for example - a PMDA that calculates all RPC > metrics of 5 hosts it monitors, therefore inside the 'total-RPC' PMDA > there's a call to pmFetch for the rpc metrics (from linux PMDA) from the > other hosts (the context is - PM_CONTEXT_HOST). Actually the call to > 'pmFetch' for the accumulative metrics calls pmFetch for each host - for the > calculated metrics. > Does it matter if the PMDA's are installed as DSO? > thanks in advance, gilly > From owner-pcp@oss.sgi.com Wed Jul 18 05:14:28 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6ICESr31678 for pcp-outgoing; Wed, 18 Jul 2001 05:14:28 -0700 Received: from mailgw3.netvision.net.il (mailgw3.netvision.net.il [194.90.1.11]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6ICEPV31673 for ; Wed, 18 Jul 2001 05:14:26 -0700 Received: from mail.exanet.co.il ([212.143.73.99]) by mailgw3.netvision.net.il (8.9.3/8.9.3) with ESMTP id PAA10742 for ; Wed, 18 Jul 2001 15:12:32 +0300 (IDT) Received: from cat (fw1.exanet.co.il [212.143.73.98]) by mail.exanet.co.il (8.11.0/8.11.0) with SMTP id f6IC9fg27897 for ; Wed, 18 Jul 2001 15:09:41 +0300 Reply-To: From: "gilly" To: Subject: CORRECTION: a PMDA using another PMDA Date: Wed, 18 Jul 2001 15:12:42 +0200 Message-ID: <009401c10f8b$54cb6090$2a04000a@exanet> MIME-Version: 1.0 Content-Type: text/plain; charset="windows-1255" Content-Transfer-Encoding: 7bit X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook CWS, Build 9.0.2416 (9.0.2910.0) X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400 Importance: Normal Sender: owner-pcp@oss.sgi.com Precedence: bulk Please ignore my last mail, the problem I had described was irrelevant. The thing is I don't intend to use pmie (as everyone suggested) 'cause I don't want the 'total' agent to monitor periodically the other agent - I want it to calculate the aggregated metric every time it is requested (therefore it calls pmFetch sequentially - once for each host). I'm not sure I understand completely the use of contexts and how I should implement it in my case. Any help in this subject will be appreciated, thanks. >All the answers mentioned pmie, but what I'm trying to implement (and I'm >still not sure I can do it) is to have a PMDA that performs sequential >pmFetch calls - each for a different host, and sums the results as >'total-results'. Trying to run pminfo on this 'total' PMDA I get the >following error: >the metrics' values ARE calculated correctly, but when trying to get the >instances' names (using pmGetInDom() in lookup()) - I get an "Unknown or >illegal instance identifier" error. It seems that the second call >(pmGetInDom, after pmFetch) to the PMDA on the other host doesn't reach its >destination. >Maybe I have a context problem? (I'm not sure I understand the contexts >issue well.) >Thanks, gilly From owner-pcp@oss.sgi.com Wed Jul 18 22:03:52 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6J53qL30124 for pcp-outgoing; Wed, 18 Jul 2001 22:03:52 -0700 Received: from rigel.cis.ksu.edu (root@rigel.cis.ksu.edu [129.130.10.65]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6J53oV30120 for ; Wed, 18 Jul 2001 22:03:50 -0700 Received: from pollux.cis.ksu.edu (mst9696@pollux.cis.ksu.edu [129.130.10.33]) by rigel.cis.ksu.edu (8.9.1/8.9.1/000517) with ESMTP id AAA29573; Thu, 19 Jul 2001 00:03:48 -0500 (CDT) Received: from localhost (mst9696@localhost) by pollux.cis.ksu.edu (8.9.1/8.9.1/000517) with ESMTP id AAA14749; Thu, 19 Jul 2001 00:03:47 -0500 (CDT) X-Authentication-Warning: pollux.cis.ksu.edu: mst9696 owned process doing -bs Date: Thu, 19 Jul 2001 00:03:47 -0500 (CDT) From: Madhu Sudhan R Tera To: kenmcd@melbourne.sgi.com cc: pcp@oss.sgi.com Subject: Re: getting the currently running process metrics In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pcp@oss.sgi.com Precedence: bulk Hello ! There is no metric that gives the per process percentage of CPU cycles used.The "pminfo -dT proc.*" command (ofcourse not *).But i didnt find the one i required.Can anyone suggest me a solution. Thanks in advance.. Thank you, With regards, MadhuSudhan Reddy Tera -------------------------------------------------- |Department of Computing and Information Sciences, | |Kansas State University, | |Manhattan, KS 66502. | |ph no:785-537-3276 | -------------------------------------------------- On Wed, 11 Jul 2001 kenmcd@melbourne.sgi.com wrote: > On Fri, 6 Jul 2001, Madhu Sudhan R Tera wrote: > > > Hello! > > Thank you for your quick response. > > > > As mentioned in my previous mail,i have recently installed PCP. > > > > But is there a procedure to pass process ids and get the metrics > > while writing a PMAPI?? > > You'll need pmLookupDesc(3) to get the instance domain number associated > with the metrics you are interested in, and then pmGetInDom(3) and/or > pmLookupInDom(3) to process the instance domain that encodes the > process ids. Then pmAddProfile(3) and pmDelProfile(3) to build/update a > "profile". And finally pmFetch(3) to actually do the work. > > You can see an example of all of this in the source for pmval, e.g. > the command > > $ pmval -i 920 proc.psinfo.utime > > reports the user CPU time for process id 920. > > To get some idea of the interaction between the client (pmval in this > case) and pmcd below the PMAPI, use the PDU tracing option: > > $ pmval -i 920 -D pdu proc.psinfo.utime > > > ... And what is this process.psinfo.stime and > > others in pminfo command(As far as i understand its a namespace.Correct me > > if i am wrong!) > > I'm not going to explain them, that's the PMDA writer's job ... the -T option to pminfo requests the > help text for the named metrics, e.g. > > $ pminfo -T proc.psinfo.utime > > proc.psinfo.utime > Help: > time (in ms) spent executing user code since process started > > From owner-pcp@oss.sgi.com Sun Jul 29 16:37:53 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6TNbrQ29922 for pcp-outgoing; Sun, 29 Jul 2001 16:37:53 -0700 Received: from deliverator.sgi.com (deliverator.sgi.com [204.94.214.10]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6TNbpV29913 for ; Sun, 29 Jul 2001 16:37:51 -0700 Received: from rattle.melbourne.sgi.com (rattle.melbourne.sgi.com [134.14.55.145]) by deliverator.sgi.com (980309.SGI.8.8.8-aspam-6.2/980310.SGI-aspam) via ESMTP id QAA25301 for ; Sun, 29 Jul 2001 16:37:36 -0700 (PDT) mail_from (kenmcd@melbourne.sgi.com) Received: from localhost (kenmcd@localhost) by rattle.melbourne.sgi.com (SGI-8.9.3/8.9.3) with ESMTP id JAA62062; Mon, 30 Jul 2001 09:36:29 +1000 (AEST) X-Authentication-Warning: rattle.melbourne.sgi.com: kenmcd owned process doing -bs Date: Mon, 30 Jul 2001 09:36:29 +1000 From: Ken McDonell Reply-To: To: cc: Subject: Re: linux pmda Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pcp@oss.sgi.com Precedence: bulk [this bounced off the list sever due to a very long attachment] ---------- Forwarded message ---------- From: "gilly" To: Subject: linux pmda Date: Thu, 26 Jul 2001 20:31:39 +0200 Hi 1. There might be a bug in the linux pmda concerning the cpu utilization statistics. In the pmda.c file the calculations for kernel.all.cpu.* metrics are calculated 'dangerously', the 'user' stat for example - "atom->ul = 1000 * proc_stat.user / proc_stat.hz" - multiplying by 1000 might be bigger than u_int and the result is wrong. attached a patch. 2. How can I add a new metric to the linux pmda? I would like to add 'uptime' metric (under kernel.all) to show the time the system is running. I'm not sure how to add this metric to the linux pmda, it seems I can't use the pmnsadd / pmnsmerge / pmnsdel utilities because 'root_linux' is not in a fit format. Am I supposed to build a new default PMNS? thanks in advance, gilly ---------- 1. Does look like a bug and Mark Goodwin will investigate as time permits. The only good thing is that when the premature overflow does occur you get one bogus value, and there after the _rate_ of change is probably goinf to be correct, at least until the next premature overflow. 2. The sequence is something like this ... - choose a name for the new metric - define the semantics of the new metric - choose a PMID for the new metric - construct help text for the new metric now in the src/pmdas/linux directory - update root_linux to include the new metric's name and PMID - look in pmda.c + add the metric to metrictab[] + in linux_fetchCallBack() use the PMID and add code at the point where the PMID components are decoded to instantiate the value of the new metric + skip the instance domain stuff, this metric will be singular - add the help text to "help" - make install now in the /var/pcp/pmns directory, run # ./Rebuild -u and restart pmcd. From owner-pcp@oss.sgi.com Mon Jul 30 20:06:58 2001 Received: (from majordomo@localhost) by oss.sgi.com (8.11.2/8.11.3) id f6V36wk25410 for pcp-outgoing; Mon, 30 Jul 2001 20:06:58 -0700 Received: from deliverator.sgi.com (deliverator.sgi.com [204.94.214.10]) by oss.sgi.com (8.11.2/8.11.3) with SMTP id f6V36rV25402 for ; Mon, 30 Jul 2001 20:06:53 -0700 Received: from larry.melbourne.sgi.com (larry.melbourne.sgi.com [134.14.52.130]) by deliverator.sgi.com (980309.SGI.8.8.8-aspam-6.2/980310.SGI-aspam) via SMTP id UAA14810 for ; Mon, 30 Jul 2001 20:06:39 -0700 (PDT) mail_from (markgw@sgi.com) Received: from sherman.melbourne.sgi.com (sherman.melbourne.sgi.com [134.14.55.175]) by larry.melbourne.sgi.com (950413.SGI.8.6.12/950213.SGI.AUTOCF) via ESMTP id NAA22942; Tue, 31 Jul 2001 13:05:30 +1000 Date: Tue, 31 Jul 2001 13:05:30 +1000 (EST) From: Mark Goodwin X-Sender: markgw@sherman.melbourne.sgi.com To: kenmcd@sgi.com cc: gilly@exanet.com, pcp@oss.sgi.com Subject: Re: linux pmda In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: MULTIPART/MIXED; BOUNDARY="243707703-729795391-996548730=:23819" Sender: owner-pcp@oss.sgi.com Precedence: bulk This message is in MIME format. The first part should be readable text, while the remaining parts are likely unreadable without MIME-aware tools. Send mail to mime@docserver.cac.washington.edu for more info. --243707703-729795391-996548730=:23819 Content-Type: TEXT/PLAIN; charset=US-ASCII On Mon, 30 Jul 2001, Ken McDonell wrote: > From: "gilly" > To: > Subject: linux pmda > Date: Thu, 26 Jul 2001 20:31:39 +0200 > > Hi > 1. There might be a bug in the linux pmda concerning the cpu utilization > statistics. > In the pmda.c file the calculations for kernel.all.cpu.* metrics are > calculated 'dangerously', the 'user' stat for example - > "atom->ul = 1000 * proc_stat.user / proc_stat.hz" - multiplying by 1000 > might be bigger than u_int and the result is wrong. > attached a patch. > > thanks in advance, > gilly > ---------- > > 1. Does look like a bug and Mark Goodwin will investigate as time permits. > The only good thing is that when the premature overflow does occur you > get one bogus value, and there after the _rate_ of change is probably > goinf to be correct, at least until the next premature overflow. Gilly, I have looked at your proposed fix, and applied a similar change to that proposed for the kernel.all.cpu.* and kernel.percpu.cpu.* metrics, as follows: Original (overflow problems) : atom->ul = 1000 * proc_stat.user / proc_stat.hz; Your proposed fix (avoids overflow, but looses precision) : atom->ul = (double)((proc_stat.user / proc_stat.hz)) * 1000; Final fix: (avoids overflow and preserves precision) : atom->ul = 1000 * (double)proc_stat.user / proc_stat.hz; > 2. How can I add a new metric to the linux pmda? > I would like to add 'uptime' metric (under kernel.all) to show the time the > system is running. > I'm not sure how to add this metric to the linux pmda, it seems I can't use > the pmnsadd / pmnsmerge / pmnsdel utilities because 'root_linux' is not in a > fit format. > Am I supposed to build a new default PMNS? > Ken already answered your general questions. In any case, I noticed the file you attached had some of the work done for the new metric kernel.all.uptime, so I finished it off, with due acknowledgement. The new metric is a count of the number of hours since boot, with instantaneous semantics (i.e. not counter). sherman 46% pminfo -h sandpit -fmdt kernel.all.uptime kernel.all.uptime PMID: 60.0.49 [time the current kernel has been running] Data Type: 32-bit unsigned int InDom: PM_INDOM_NULL 0xffffffff Semantics: instant Units: hour count value 356 Does this sound reasonable? Would people prefer minutes or seconds? Or would that only be relevant for a Windows PMDA? ;-) Which reminds me, was someone attempting a win32/cygwin port? Any progress? thanks -- Mark ps: in the future, please send patches like the one attached rather than whole files. The latter overflows the mail server at oss.sgi.com, resulting in bounced meesages. --243707703-729795391-996548730=:23819 Content-Type: TEXT/PLAIN; charset=US-ASCII; name=patch Content-Transfer-Encoding: BASE64 Content-ID: Content-Description: Content-Disposition: attachment; filename=patch DQo9PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT0NCkNIQU5HRUxPRw0K PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09DQoNCi0tLSAvdXNyL3Rt cC9UbXBEaXIuODk0OS0wL0NIQU5HRUxPR18xLjg1CVR1ZSBKdWwgMzEgMTM6 MDI6MDUgMjAwMQ0KKysrIENIQU5HRUxPRwlUdWUgSnVsIDMxIDEyOjEwOjA2 IDIwMDENCkBAIC0yNjQsMyArMjY0LDYgQEANCiAgICAgLSBmaXggY29uZmln dXJlLmluIHRvIHdvcmsgd2hlbiBwcyBnaXZlcyB3YXJuaW5ncyBvbiBzdGRl cnI7DQogICAgICAgKGVnLiBTeXN0ZW0ubWFwIGRvZXNuJ3QgbWF0Y2ggcnVu bmluZyBrZXJuZWwpDQogICAgIC0gYWRkZWQgbG1zZW5zb3JzIFBNREEsIGNv bnRyaWJ1dGVkIGJ5IFRyb3kgRGF3c29uIDxkYXdzb25AZm5hbC5nb3Y+DQor ICAgIC0gYWRkZWQga2VybmVsLmFsbC51cHRpbWUsIGNvbnRyaWJ1dGVkIGJ5 IEdpbGx5IDxnaWxseUBleGFuZXQuY29tPg0KKyAgICAtIGZpeGVkIGludCBv dmVyZmxvdyB3aXRoIGtlcm5lbC4qLmNwdS4qIG1ldHJpY3MuIEZpeCBjb250 cmlidXRlZA0KKyAgICAgIGJ5IEdpbGx5IDxnaWxseUBleGFuZXQuY29tPg0K DQo9PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT0NClZFUlNJT04ucGNw DQo9PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT0NCg0KLS0tIC91c3Iv dG1wL1RtcERpci44OTQ5LTAvVkVSU0lPTi5wY3BfMS42MAlUdWUgSnVsIDMx IDEzOjAyOjA1IDIwMDENCisrKyBWRVJTSU9OLnBjcAlUdWUgSnVsIDMxIDEy OjA4OjU3IDIwMDENCkBAIC01LDcgKzUsNyBAQA0KIFBBQ0tBR0VfTUFKT1I9 Mg0KIFBBQ0tBR0VfTUlOT1I9Mg0KIFBBQ0tBR0VfUkVWSVNJT049Mg0KLVBB Q0tBR0VfQlVJTEQ9Mg0KK1BBQ0tBR0VfQlVJTEQ9Mw0KIFBBQ0tBR0VfVkVS U0lPTj0ke1BBQ0tBR0VfTUFKT1J9LiR7UEFDS0FHRV9NSU5PUn0uJHtQQUNL QUdFX1JFVklTSU9OfQ0KIFBBQ0tBR0VfRElTVFJJQlVUSU9OPSJTR0kgT3Bl biBTb3VyY2UiDQogDQoNCj09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PQ0Kc3JjL3BtZGFzL2xpbnV4L2hlbHANCj09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PQ0KDQotLS0gL3Vzci90bXAvVG1wRGlyLjg5NDktMC9zcmMv cG1kYXMvbGludXgvaGVscF8xLjE4CVR1ZSBKdWwgMzEgMTM6MDI6MDUgMjAw MQ0KKysrIHNyYy9wbWRhcy9saW51eC9oZWxwCVR1ZSBKdWwgMzEgMTE6NTM6 MTYgMjAwMQ0KQEAgLTg2NCw3ICs4NjQsNyBAQA0KIEAgaGludi5jcHUuY2Fj aGUgcHJpbWFyeSBjYWNoZSBzaXplIG9mIGVhY2ggQ1BVIGFzIHJlcG9ydGVk IGJ5IC9wcm9jL2NwdWluZm8NCiBAIGhpbnYuY3B1LmJvZ29taXBzIGJvZ28g bWlwcyByYXRpbmcgZm9yIGVhY2ggQ1BVIGFzIHJlcG9ydGVkIGJ5IC9wcm9j L2NwdWluZm8NCiBAIGtlcm5lbC5hbGwuaHogdmFsdWUgb2YgSFogKGppZmZp ZXMvc2Vjb25kKSBmb3IgdGhlIGN1cnJlbnRseSBydW5uaW5nIGtlcm5lbA0K LQ0KK0Aga2VybmVsLmFsbC51cHRpbWUgdGltZSB0aGUgY3VycmVudCBrZXJu ZWwgaGFzIGJlZW4gcnVubmluZw0KIEAgbWVtLnNsYWJpbmZvLm9iamVjdHMu YWN0aXZlIG51bWJlciBvZiBhY3RpdmUgb2JqZWN0cyBpbiBlYWNoIGNhY2hl DQogQCBtZW0uc2xhYmluZm8ub2JqZWN0cy50b3RhbCB0b3RhbCBudW1iZXIg b2Ygb2JqZWN0cyBpbiBlYWNoIGNhY2hlDQogQCBtZW0uc2xhYmluZm8ub2Jq ZWN0cy5zaXplIHNpemUgb2YgaW5kaXZpZHVhbCBvYmplY3RzIG9mIGVhY2gg Y2FjaGUNCg0KPT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09DQpzcmMv cG1kYXMvbGludXgvcG1kYS5jDQo9PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT0NCg0KLS0tIC91c3IvdG1wL1RtcERpci44OTQ5LTAvc3JjL3BtZGFz L2xpbnV4L3BtZGEuY18xLjMxCVR1ZSBKdWwgMzEgMTM6MDI6MDUgMjAwMQ0K KysrIHNyYy9wbWRhcy9saW51eC9wbWRhLmMJVHVlIEp1bCAzMSAxMjo1OToy OSAyMDAxDQpAQCAtNDIwLDEwICs0MjAsMTUgQEANCiAgICAgICBQTURBX1BN VU5JVFMoMCwxLDAsMCxQTV9USU1FX01TRUMsMCkgfSwgfSwNCiANCiAvKiBr ZXJuZWwuYWxsLmh6ICovDQotICAgIHsgJnByb2Nfc3RhdC5oeiwgDQorICAg IHsgTlVMTCwNCiAgICAgICB7IFBNREFfUE1JRChDTFVTVEVSX1NUQVQsNDgp LCBQTV9UWVBFX1UzMiwgUE1fSU5ET01fTlVMTCwgUE1fU0VNX0RJU0NSRVRF LCANCiAgICAgICBQTURBX1BNVU5JVFMoMCwtMSwxLDAsUE1fVElNRV9TRUMs UE1fQ09VTlRfT05FKSB9LCB9LA0KIA0KKy8qIGtlcm5lbC5hbGwudXB0aW1l ICovDQorICAgIHsgTlVMTCwNCisgICAgICB7IFBNREFfUE1JRChDTFVTVEVS X1NUQVQsNDkpLCBQTV9UWVBFX1UzMiwgUE1fSU5ET01fTlVMTCwgUE1fU0VN X0lOU1RBTlQsIA0KKyAgICAgIFBNREFfUE1VTklUUygwLDEsMSwwLFBNX1RJ TUVfSE9VUixQTV9DT1VOVF9PTkUpIH0sIH0sDQorDQogDQogLyoNCiAgKiAv cHJvYy9tZW1pbmZvIGNsdXN0ZXINCkBAIC0yNTA1LDE2ICsyNTEwLDE2IEBA DQogCSAgICAgKi8NCiAJICAgIHN3aXRjaCAoaWRwLT5pdGVtKSB7DQogCSAg ICBjYXNlIDA6IC8qIHVzZXIgKi8NCi0JCWF0b20tPnVsID0gMTAwMCAqIHBy b2Nfc3RhdC5wX3VzZXJbaW5zdF0gLyBwcm9jX3N0YXQuaHo7DQorCQlhdG9t LT51bCA9IDEwMDAgKiAoZG91YmxlKXByb2Nfc3RhdC5wX3VzZXJbaW5zdF0g LyBwcm9jX3N0YXQuaHo7DQogCQlicmVhazsNCiAJICAgIGNhc2UgMTogLyog bmljZSAqLw0KLQkJYXRvbS0+dWwgPSAxMDAwICogcHJvY19zdGF0LnBfbmlj ZVtpbnN0XSAvIHByb2Nfc3RhdC5oejsNCisJCWF0b20tPnVsID0gMTAwMCAq IChkb3VibGUpcHJvY19zdGF0LnBfbmljZVtpbnN0XSAvIHByb2Nfc3RhdC5o ejsNCiAJCWJyZWFrOw0KIAkgICAgY2FzZSAyOiAvKiBzeXMgKi8NCi0JCWF0 b20tPnVsID0gMTAwMCAqIHByb2Nfc3RhdC5wX3N5c1tpbnN0XSAvIHByb2Nf c3RhdC5oejsNCisJCWF0b20tPnVsID0gMTAwMCAqIChkb3VibGUpcHJvY19z dGF0LnBfc3lzW2luc3RdIC8gcHJvY19zdGF0Lmh6Ow0KIAkJYnJlYWs7DQog CSAgICBjYXNlIDM6IC8qIGlkbGUgKi8NCi0JCWF0b20tPnVsID0gMTAwMCAq IHByb2Nfc3RhdC5wX2lkbGVbaW5zdF0gLyBwcm9jX3N0YXQuaHo7DQorCQlh dG9tLT51bCA9IDEwMDAgKiAoZG91YmxlKXByb2Nfc3RhdC5wX2lkbGVbaW5z dF0gLyBwcm9jX3N0YXQuaHo7DQogCQlicmVhazsNCiANCiAJICAgIGNhc2Ug NDogLyogcmlvICovDQpAQCAtMjU1MiwxNyArMjU1NywxOCBAQA0KIAkJYXRv bS0+dWwgPSBwcm9jX3N0YXQucHJvY2Vzc2VzOw0KIAkJYnJlYWs7DQogDQor CSAgICAvKiBnaWxseSAtIGNoYW5nZSB0aGUgY2FsY3VsYXRpb24gdG8gcHJl dmVudCBhIGJ1ZyAqLw0KIAkgICAgY2FzZSAyMDogLyogYWxsLnVzZXIgKi8N Ci0JCWF0b20tPnVsID0gMTAwMCAqIHByb2Nfc3RhdC51c2VyIC8gcHJvY19z dGF0Lmh6Ow0KKwkJYXRvbS0+dWwgPSAxMDAwICogKGRvdWJsZSlwcm9jX3N0 YXQudXNlciAvIHByb2Nfc3RhdC5oejsNCiAJCWJyZWFrOw0KIAkgICAgY2Fz ZSAyMTogLyogYWxsLm5pY2UgKi8NCi0JCWF0b20tPnVsID0gMTAwMCAqIHBy b2Nfc3RhdC5uaWNlIC8gcHJvY19zdGF0Lmh6Ow0KKwkJYXRvbS0+dWwgPSAx MDAwICogKGRvdWJsZSlwcm9jX3N0YXQubmljZSAvIHByb2Nfc3RhdC5oejsN CiAJCWJyZWFrOw0KIAkgICAgY2FzZSAyMjogLyogYWxsLnN5cyAqLw0KLQkJ YXRvbS0+dWwgPSAxMDAwICogcHJvY19zdGF0LnN5cyAvIHByb2Nfc3RhdC5o ejsNCisJCWF0b20tPnVsID0gMTAwMCAqIChkb3VibGUpcHJvY19zdGF0LnN5 cyAvIHByb2Nfc3RhdC5oejsNCiAJCWJyZWFrOw0KIAkgICAgY2FzZSAyMzog LyogYWxsLmlkbGUgKi8NCi0JCWF0b20tPnVsID0gMTAwMCAqIHByb2Nfc3Rh dC5pZGxlIC8gcHJvY19zdGF0Lmh6Ow0KKwkJYXRvbS0+dWwgPSAxMDAwICog KGRvdWJsZSlwcm9jX3N0YXQuaWRsZSAvIHByb2Nfc3RhdC5oejsNCiAJCWJy ZWFrOw0KIA0KIAkgICAgY2FzZSAyNDogLyogYWxsLnJpbyAqLw0KQEAgLTI2 MjgsNiArMjYzNCwxOCBAQA0KIAkgICAgY2FzZSA0NjogLyogZGlzay5kZXYu YXZhY3RpdmUgKi8NCiAJICAgIGNhc2UgNDc6IC8qIGRpc2suZGV2LmF2ZXEg Ki8NCiAJCXJldHVybiAwOyAvKiBubyB2YWx1ZXMgYXZhaWxhYmxlICovDQor DQorCSAgICBjYXNlIDQ4OiAvKiBrZXJuZWwuYWxsLmh6ICovDQorCSAgICAJ YXRvbS0+dWwgPSBwcm9jX3N0YXQuaHo7DQorCQlicmVhazsNCisNCisJICAg IGNhc2UgNDk6DQorCSAgICAJLyoNCisJCSAqIGtlcm5lbC5hbGwudXB0aW1l IGluIGhvdXJzDQorCQkgKiBjb250cmlidXRlZCBieSAiZ2lsbHkiIDxnaWxs eUBleGFuZXQuY29tPg0KKwkJICovDQorCQlhdG9tLT51bCA9ICh1X2xvbmcp KHRpbWVzKE5VTEwpKSAvIChwcm9jX3N0YXQuaHogKiA2MCAqIDYwKTsNCisJ CWJyZWFrOw0KIA0KIAkgICAgZGVmYXVsdDoNCiAJCXJldHVybiBQTV9FUlJf UE1JRDsNCg0KPT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09DQpzcmMv cG1kYXMvbGludXgvcm9vdF9saW51eA0KPT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09DQoNCi0tLSAvdXNyL3RtcC9UbXBEaXIuODk0OS0wL3NyYy9w bWRhcy9saW51eC9yb290X2xpbnV4XzEuMTYJVHVlIEp1bCAzMSAxMzowMjow NSAyMDAxDQorKysgc3JjL3BtZGFzL2xpbnV4L3Jvb3RfbGludXgJVHVlIEp1 bCAzMSAxMTo1MTo0MiAyMDAxDQpAQCAtNTUsNiArNTUsNyBAQA0KICAgICBz eXNmb3JrCQk2MDowOjE0DQogICAgIHN5c2NhbGwJCTYwOjQ6MQ0KICAgICBo egkJCTYwOjA6NDgNCisgICAgdXB0aW1lCQk2MDowOjQ5DQogfQ0KIA0KIGtl cm5lbC5hbGwuY3B1IHsNCg== --243707703-729795391-996548730=:23819--