pcp
[Top] [All Lists]

pcp updates

To: pcp@xxxxxxxxxxx
Subject: pcp updates
From: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Date: Mon, 22 Apr 2013 15:38:46 +1000
Delivered-to: pcp@xxxxxxxxxxx
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:17.0) Gecko/20130329 Thunderbird/17.0.5
Mostly QA gopher smacking, but there is one small libpcp mem leak plugger in 
this lot.

Changes committed to git://oss.sgi.com/kenj/pcp.git dev

 qa/066                     |   11 --
 qa/066.out.5               |    6 -
 qa/087                     |    2 
 qa/087.out                 |    6 -
 qa/200                     |    4 
 qa/200.out                 |   61 ------------
 qa/200.out.2               |   23 ----
 qa/287                     |   10 --
 qa/308                     |   51 +++++++++-
 qa/369                     |   23 +++-
 qa/369.out.1               |  172 +++++++++++++++++-----------------
 qa/369.out.2               |  224 ++++++++++++++++++++++-----------------------
 qa/513                     |   12 +-
 qa/513.out                 |    9 +
 qa/admin/check-vm          |    2 
 qa/admin/show-me-all       |    9 +
 qa/group                   |    2 
 qa/src/GNUmakefile         |    2 
 src/libpcp/src/p_profile.c |    6 -
 19 files changed, 312 insertions(+), 323 deletions(-)

commit 5901555c7a6d67331311f4fbe3681983b5ce2509
Author: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Date:   Mon Apr 22 15:37:23 2013 +1000

    qa/admin/show-me-all - report hostname

commit 096726c87cda61784c47619192b810d527e1eb96
Author: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Date:   Sun Apr 21 19:55:18 2013 +1000

    qa/group - test 308 uses the proc pmda

commit f9ef0b810c49d90905fdc746ab6a5a7334d52f35
Author: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Date:   Sun Apr 21 19:54:07 2013 +1000

    libpcp/p_profile.c - plug small memory leak
    
    Exposed when qa/513 was fixed.

commit c4c808c612b7a75f40d42bdbb45211398eb822b9
Author: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Date:   Sun Apr 21 19:51:13 2013 +1000

    qa/513 - use "proper" valgrind infrastructure
    
    This test was using a feral valgrind process.  Change to use the
    "standard" process, namely:
    
        _check_valgrind
        ...
        _run_valgrind <some command>

commit 3c7c82980a3cc2154c44e82495e315e2b5bffe14
Author: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Date:   Sun Apr 21 17:02:17 2013 +1000

    qa/308 - more broken than I first thought
    
    Needs the proc PMDA installed locally and on 2 remote machines.
    
    Make it so locally, clean up when you're done and harden the test
    with better integrity checks for the expected instance in the proc
    pid indom.

commit 160dcaed0a88e66b5afd883ee1dfc3026d301bf0
Author: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Date:   Sun Apr 21 16:41:28 2013 +1000

    qa/308 - accommodate Fedora 18 and proc PMDA changes
    
    In Fedora 18, the process with pid 1 is not "init" but
    /usr/lib/systemd/systemd (with some complicated arguments).
    
    Also the PMID we're filtering has probably been wrong since the
    proc metrics were split into a separate PMDA.
    
    Only started to fail when a couple of QA hosts had the proc PMDA
    installed by default, and this moved from a "notrun" to "failing
    most places" state.

commit 7d9b3c8f381d1510afefe590bd7beca6b16c198c
Author: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Date:   Sun Apr 21 07:21:22 2013 +1000

    qa/066 - ipv6 wildcard changes

commit ec72addbd4053752737ef4616ff03e2b1cbf26a9
Author: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Date:   Sun Apr 21 07:15:25 2013 +1000

    qa/GNUmakefile - localconfig.h depends on ../localconfig

commit 81a7cbb97afe4c28d0abe32af653e0e85bbb8bff
Author: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Date:   Sun Apr 21 06:31:20 2013 +1000

    qa/admin - fiddling about .. more packages to check, cosmetics

commit 1edd70e03bd3cdf14f9a8d5ca0d64fe0360e9934
Author: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Date:   Sun Apr 21 06:29:47 2013 +1000

    qa/087 - relax one of the expected number of log reads ranges
    
    Seems you can be lucky in creating the archive, and sometimes
    require a couple of less log reads for one of the test cases.

commit 10c795732a75e9b631150ebb110c3dfa0bf8c202
Author: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Date:   Sun Apr 21 06:27:00 2013 +1000

    qa/200 - a couple of changes
    
    1. rm 200.out that was missed when the expected output was split into
       200.out.1 and 200.out.2
    2. Filter out the (recently added) libpcp diagnostic
        __pmConnect(fd=..., 127.0.0.1) -> ...
       as the status reported is not deterministic and does not
       change the correctness of this test

commit 25df3783de9868248f2d5df5b9cea7dacb77297d
Author: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Date:   Sun Apr 21 06:24:09 2013 +1000

    qa/369 - sort symbol table dump in output

commit 0b50ff773f613e8194af7e25066b9f84508d8204
Author: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Date:   Sun Apr 21 06:22:53 2013 +1000

    qa/287 - uname -m more reliable than uname -p

commit c6594ea311e3225ef5df546b4b75f8faab3a8c2a
Author: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Date:   Sun Apr 21 06:08:08 2013 +1000

    qa/066 - more helpful notrun message

<Prev in Thread] Current Thread [Next in Thread>