Home - Waterfall Grid T-Grid Console Builders Recent Builds Buildslaves Changesources - JSON API - About

Builder epel-7-x86_64 Build #308

Results:

Build successful

SourceStamp:

Repositoryhttps://github.com/ClusterLabs/pacemaker
Branchmaster
Revision0ea7a77e12094f7410870a96be5c635f8eb272cf
Got Revision0ea7a77e12094f7410870a96be5c635f8eb272cf
Changes17 changes

BuildSlave:

build-uk

Reason:

downstream

Steps and Logfiles:

  1. git update ( 9 secs )
    1. stdio
  2. setproperty property 'buildroot' set ( 0 secs )
    1. stdio
    2. property changes
  3. datetime property 'datetime' set ( 0 secs )
    1. stdio
    2. property changes
  4. destdir 2 properties set ( 0 secs )
    1. stdio
    2. property changes
  5. mkdir Ran ( 0 secs )
    1. stdio
  6. branch_perms Ran ( 0 secs )
    1. stdio
  7. download downloading to mock.cfg ( 0 secs )
    1. - no logs -
  8. defaults 'cp /etc/mock/site-defaults.cfg ...' ( 0 secs )
    1. stdio
  9. srpm property 'srpm_file' set ( 2 secs )
    1. stdio
    2. property changes
  10. build Mock Build ( 4 mins, 51 secs )
    1. stdio
    2. Setup
    3. Build
    4. State
  11. cleanup Cleanup ( 0 secs )
    1. stdio
  12. rpmlint Finished checking RPM/SPEC issues ( 7 secs )
    1. stdio
    2. 3 Warnings
  13. install Package install ( 8 secs )
    1. stdio
  14. basic_sanity Regressions ( 10 mins, 26 secs )
    1. stdio
  15. valgrind_sanity Valgrind Regressions skipped ( 0 secs )
    1. - no logs -
  16. lcov Coverage Analysis skipped ( 0 secs )
    1. - no logs -
  17. lcov_extract Extract Coverage Results skipped ( 0 secs )
    1. - no logs -
  18. lcov_publish lcov_publish skipped ( 0 secs )
    1. - no logs -
  19. docker Docker skipped ( 0 secs )
    1. - no logs -
  20. copy_pkg uploading mock ( 15 secs )
    1. - no logs -
  21. purge_old Ran ( 0 secs )
    1. stdio
  22. www_perms Ran ( 0 secs )
    1. stdio

Build Properties:

NameValueSource
branch master Build
builddir /home/buildbot/slave/pacemaker/epel-7-x86_64 Slave
buildername epel-7-x86_64 Builder
buildnumber 308 Build
buildroot /home/buildbot/slave/pacemaker/epel-7-x86_64/build SetProperty Step
codebase Build
datetime 2017-04-18-20:15:37 SetProperty Step
destdir /var/www/html/buildbot/master/epel-7-x86_64/308-2017-04-18-20:15:37 SetProperty Step
destwww buildbot/master/epel-7-x86_64/308-2017-04-18-20:15:37 SetProperty Step
got_revision 0ea7a77e12094f7410870a96be5c635f8eb272cf Git
project Build
repository https://github.com/ClusterLabs/pacemaker Build
revision 0ea7a77e12094f7410870a96be5c635f8eb272cf Build
scheduler Pacemaker master Scheduler
slavename build-uk BuildSlave
srpm_file /home/buildbot/slave/pacemaker/epel-7-x86_64/build/pacemaker-1.1.17-0.308.0ea7a77.git.epel.src.rpm SetProperty Step

Forced Build Properties:

NameLabelValue

Responsible Users:

  1. Ken Gaillot

Timing:

StartTue Apr 18 22:15:28 2017
EndTue Apr 18 22:31:29 2017
Elapsed16 mins, 0 secs

All Changes:

:

  1. Change #13267

    Category None
    Changed by Ken Gaillot <kgaillotohnoyoudont@redhat.com>
    Changed at Tue 11 Apr 2017 00:06:50
    Repository https://github.com/ClusterLabs/pacemaker
    Branch master
    Revision b005b4f2809020304862000326b22cded7b14377

    Comments

    Fix: libpe_status: guest nodes from bundles may have attributes
    
    Previously, if a guest node created by a bundle had a node attribute,
    pe_create_node() would get called twice, once when parsing the <node> entry and
    once when parsing the <bundle>, resulting in any bundle primitive being unable
    to run.

    Changed files

    • lib/pengine/container.c
  2. Change #13268

    Category None
    Changed by Ken Gaillot <kgaillotohnoyoudont@redhat.com>
    Changed at Tue 11 Apr 2017 00:06:50
    Repository https://github.com/ClusterLabs/pacemaker
    Branch master
    Revision 7b89ff8b65fcdcad55676578361080eb23edb3e4

    Comments

    Low: fencing: ignore empty 'action' parameter in fence devices
    
    This makes the fix in 9c0c3d6 more comprehensive.

    Changed files

    • fencing/commands.c
  3. Change #13269

    Category None
    Changed by Ken Gaillot <kgaillotohnoyoudont@redhat.com>
    Changed at Tue 18 Apr 2017 00:23:21
    Repository https://github.com/ClusterLabs/pacemaker
    Branch master
    Revision 100dd5fda476ef526ac1964260252b30864d5ca7

    Comments

    Fix: crmd: check for too many stonith failures only when aborting for that reason
    
    Previously, crmd would check for too many stonith failures whenever aborting
    a transition. This would lead to a new transition not being triggered when
    aborting for some other unrelated reason, such as a configuration change.
    
    Now, crmd checks for too many stonith failures only when aborting due to a new
    stonith failure.

    Changed files

    • crmd/crmd_utils.h
    • crmd/te_actions.c
    • crmd/te_callbacks.c
    • crmd/te_utils.c
  4. Change #13270

    Category None
    Changed by Ken Gaillot <kgaillotohnoyoudont@redhat.com>
    Changed at Tue 18 Apr 2017 00:23:22
    Repository https://github.com/ClusterLabs/pacemaker
    Branch master
    Revision 3c49a1cf86cb819eca18c841661d90fa65bcb185

    Comments

    Low: crmd: consider target when checking stonith failures
    
    Previously, if the crmd aborted a transition due to failure to fence a
    particular node, a new transition would not be started if *any* node had
    been fenced too many times. Now, only failures of the particular target are
    checked in that situation.

    Changed files

    • crmd/crmd_utils.h
    • crmd/te_callbacks.c
    • crmd/te_utils.c
  5. Change #13271

    Category None
    Changed by Ken Gaillot <kgaillotohnoyoudont@redhat.com>
    Changed at Tue 18 Apr 2017 00:23:22
    Repository https://github.com/ClusterLabs/pacemaker
    Branch master
    Revision 0c43785dad9be38566cccce677c54da42ff2c691

    Comments

    Fix: crmd: forget stonith failures when forgetting node

    Changed files

    • crmd/messages.c
  6. Change #13272

    Category None
    Changed by Ken Gaillot <kgaillotohnoyoudont@redhat.com>
    Changed at Tue 18 Apr 2017 00:23:22
    Repository https://github.com/ClusterLabs/pacemaker
    Branch master
    Revision 515424f01b8ac5eb8705cecb26a60e17de3a7df6

    Comments

    Fix: crmd: track stonith fail counts on all nodes
    
    Previously, the stonith fail count was incremented in
    tengine_stonith_callback(), which is called only on the DC. Now, it is
    incremented in tengine_stonith_notify() instead, which is called on all nodes,
    ensuring the count is correct when a new node takes over DC.

    Changed files

    • crmd/crmd_utils.h
    • crmd/te_callbacks.c
    • crmd/te_utils.c
  7. Change #13273

    Category None
    Changed by Ken Gaillot <kgaillotohnoyoudont@redhat.com>
    Changed at Tue 18 Apr 2017 00:23:22
    Repository https://github.com/ClusterLabs/pacemaker
    Branch master
    Revision 714a8d07a500675d84e6ef779ba21e6c23e27853

    Comments

    Low: crmd: allow clearing all stonith fail counts
    
    for future improvements

    Changed files

    • crmd/te_callbacks.c
  8. Change #13274

    Category None
    Changed by Ken Gaillot <kgaillotohnoyoudont@redhat.com>
    Changed at Tue 18 Apr 2017 00:23:22
    Repository https://github.com/ClusterLabs/pacemaker
    Branch master
    Revision 8fd6691558d94a8294f3d860cc9451c1a8e0c7a1

    Comments

    Low: crmd: skip restart at (not above) stonith-max-attempts

    Changed files

    • crmd/te_callbacks.c
  9. Change #13275

    Category None
    Changed by Ken Gaillot <kgaillotohnoyoudont@redhat.com>
    Changed at Tue 18 Apr 2017 01:23:40
    Repository https://github.com/ClusterLabs/pacemaker
    Branch master
    Revision 9e9a271fd666ff371487f22c28ba9e420a22434c

    Comments

    Fix: crmd: don't restart transition if no fence devices
    
    This restores the behavior removed by ff881376, but more precisely where it's
    needed.

    Changed files

    • crmd/crmd_utils.h
    • crmd/te_callbacks.c
    • crmd/te_utils.c
    • include/crm/transition.h
  10. Change #13276

    Category None
    Changed by Ken Gaillot <kgaillotohnoyoudont@redhat.com>
    Changed at Tue 18 Apr 2017 01:23:40
    Repository https://github.com/ClusterLabs/pacemaker
    Branch master
    Revision 268f70f9ab644783a8038aa82bcca3628cc942dc

    Comments

    Low: crmd: avoid DC sending offer to itself twice

    Changed files

    • crmd/join_dc.c
  11. Change #13277

    Category None
    Changed by Ken Gaillot <kgaillotohnoyoudont@redhat.com>
    Changed at Tue 18 Apr 2017 01:23:40
    Repository https://github.com/ClusterLabs/pacemaker
    Branch master
    Revision 249c7c9a83371a67e573276a285697994fe66fed

    Comments

    Fix: crmd: don't fence old DC if it's shutting down as soon-to-be DC joins
    
    Existing peers of a DC that is shutting down can avoid fencing it (by updating
    its expected state) because it broadcasts its shutdown request. However, a
    newly joining node won't get that broadcast.
    
    Previously, if the joining node became the new DC, it would fence the old one.
    Now, the DC notifies joining nodes (via a join message field) whether it is
    shutting down, and joining nodes update its expected state accordingly.

    Changed files

    • crmd/join_client.c
    • crmd/join_dc.c
    • include/crm/msg_xml.h
  12. Change #13278

    Category None
    Changed by Ken Gaillot <kgaillotohnoyoudont@redhat.com>
    Changed at Tue 18 Apr 2017 02:16:33
    Repository https://github.com/ClusterLabs/pacemaker
    Branch master
    Revision 9fe47194b7636bfe3aebdeece0ec89a7f588d77d

    Comments

    Refactor: extra: use whitespace consistently in Dummy
    
    to make changes easier

    Changed files

    • extra/resources/Dummy
  13. Change #13279

    Category None
    Changed by Ken Gaillot <kgaillotohnoyoudont@redhat.com>
    Changed at Tue 18 Apr 2017 02:16:33
    Repository https://github.com/ClusterLabs/pacemaker
    Branch master
    Revision 14d5eb9bcf1a00612fbe952d945b07bbe3c44844

    Comments

    Test: extra: Dummy stop shouldn't fail if monitor is in progress

    Changed files

    • extra/resources/Dummy
  14. Change #13280

    Category None
    Changed by Ken Gaillot <kgaillotohnoyoudont@redhat.com>
    Changed at Tue 18 Apr 2017 02:16:33
    Repository https://github.com/ClusterLabs/pacemaker
    Branch master
    Revision 49de190ba19e28af0d4180e45cd1b585757c936a

    Comments

    Doc: Pacemaker Explained: clarify stonith-max-attempts

    Changed files

    • doc/Pacemaker_Explained/en-US/Ch-Options.txt
  15. Change #13281

    Category None
    Changed by Ken Gaillot <kgaillotohnoyoudont@redhat.com>
    Changed at Tue 18 Apr 2017 02:16:33
    Repository https://github.com/ClusterLabs/pacemaker
    Branch master
    Revision d8f99a0c06124011595b2392018ba0300505dc73

    Comments

    Fix: libcrmcommon: filter attributes with '#' from XML fields
    
    The PE passes node attributes to fence agents using hash2metafield(),
    but this would generate invalid XML syntax with per-operation fail counts,
    which contain '#'. Cluster-generated attributes aren't intended to be passed
    anyway, so just filter such names out when adding to XML.

    Changed files

    • lib/common/xml.c
    • pengine/graph.c
  16. Change #13282

    Category None
    Changed by Ken Gaillot <kgaillotohnoyoudont@redhat.com>
    Changed at Tue 18 Apr 2017 02:16:33
    Repository https://github.com/ClusterLabs/pacemaker
    Branch master
    Revision 0c15dfd61dfedbaf0d3d25c28f38b5cbf0563a21

    Comments

    Test: pengine: add test for per-op fail count

    Changed files

    • pengine/regression.sh
    • pengine/test10/per-op-failcount.dot
    • pengine/test10/per-op-failcount.exp
    • pengine/test10/per-op-failcount.scores
    • pengine/test10/per-op-failcount.summary
    • pengine/test10/per-op-failcount.xml
  17. Change #13283

    Category None
    Changed by Ken Gaillot <kgaillotohnoyoudont@redhat.com>
    Changed at Tue 18 Apr 2017 17:59:02
    Repository https://github.com/ClusterLabs/pacemaker
    Branch master
    Revision 0ea7a77e12094f7410870a96be5c635f8eb272cf

    Comments

    Merge pull request #1263 from kgaillot/fixes
    
    Better stonith failure handling in crmd, plus a few unrelated fixes

    Changed files

    • crmd/crmd_utils.h
    • crmd/join_client.c
    • crmd/join_dc.c
    • crmd/messages.c
    • crmd/te_actions.c
    • crmd/te_callbacks.c
    • crmd/te_utils.c
    • doc/Pacemaker_Explained/en-US/Ch-Options.txt
    • extra/resources/Dummy
    • fencing/commands.c
    • include/crm/msg_xml.h
    • include/crm/transition.h
    • lib/common/xml.c
    • lib/pengine/container.c
    • pengine/graph.c
    • pengine/regression.sh
    • pengine/test10/per-op-failcount.dot
    • pengine/test10/per-op-failcount.exp
    • pengine/test10/per-op-failcount.scores
    • pengine/test10/per-op-failcount.summary
    • pengine/test10/per-op-failcount.xml