Bug 599654

Summary: killing process gives CPG_REASON_LEAVE instead of CPG_REASON_PROCDOWN
Product: Red Hat Enterprise Linux 5 Reporter: David Teigland <teigland>
Component: openaisAssignee: Jan Friesse <jfriesse>
Status: CLOSED ERRATA QA Contact: Cluster QE <mspqa-list>
Severity: medium Docs Contact:
Priority: urgent    
Version: 5.5CC: cluster-maint, edamato, jkortus, sdake
Target Milestone: rcKeywords: ZStream
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: openais-0.80.6-21.el5 Doc Type: Bug Fix
Doc Text:
Previously, the Closed Process Group (CPG) interface returned the wrong result, which could have led to incorrect behavior in some situations. With this update, the CPG interface now behaves as expected.
 Story Points: ---
Clone Of:
: 604078 (view as bug list) Environment:
Last Closed: 2011-01-13 23:57:02 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Bug Depends On:    
Bug Blocks: 604078, 604242, 618765, 618766, 624488    
Attachments:
Description Flags
Proposed patch - sent to ML none

Description David Teigland 2010-06-03 17:20:08 UTC 
Description of problem:

node1: killall -9 groupd
other nodes get confchgs for all cpg's node1 was in with CPG_REASON_LEAVE

CPG_REASON_LEAVE is supposed to indicate that the process called cpg_leave().
CPG_REASON_PROCDOWN is supposed to indicate that the process exited without calling cpg_leave().

Version-Release number of selected component (if applicable):


How reproducible:


Steps to Reproduce:
1.
2.
3.
  
Actual results:


Expected results:


Additional info:
Comment 2 Jan Friesse 2010-06-15 11:24:58 UTC 
Created attachment 424119 [details]
Proposed patch - sent to ML

end CPG_REASON_PROCDOWN on process left

Our manual pages are clear:

CPG_REASON_PROCDOWN - the process left a group without calling
cpg_leave().

Currently, we are sending CPG_REASON_LEAVE in such situation.
Comment 3 Steven Dake 2010-06-27 21:37:24 UTC 
*** Bug 604242 has been marked as a duplicate of this bug. ***
Comment 4 David Teigland 2010-07-01 19:24:35 UTC 
I tested openais-0.80.6-21.el5.x86_64.rpm and it works correctly.

on nodes 1,2,3:
service cman start; service clvmd start; mount /dev/shared/x /gfs

on node 1:
killall -9 groupd

on nodes 2,3:
group_tool -v
type             level name     id       state node id local_done
fence            0     default  00010001 FAIL_START_WAIT 1 100020003 0
[2 3]
dlm              1     clvmd    00020001 FAIL_ALL_STOPPED 1 100020003 -1
[1 2 3]
dlm              1     x        00040001 FAIL_ALL_STOPPED 1 100020003 -1
[1 2 3]
gfs              2     x        00030001 FAIL_ALL_STOPPED 1 100020003 -1
[1 2 3]

They are correctly processing a node 1 failure.

Also, 'group_tool dump' data shows the node_leave event before fix and the node_down event after.

before:
1276726614 groupd confchg total 2 left 1 joined 0
1276726614 0:default confchg left 1 joined 0 total 2
1276726614 0:default confchg removed node 1 reason 2
1276726614 0:default process_node_leave 1
1276726614 0:default cpg del node 1 total 2
1276726614 0:default make_event_id 100020002 nodeid 1 memb_count 2 type 2
1276726614 0:default queue leave event for nodeid 1

after:
1278010518 groupd confchg total 2 left 1 joined 0
1278010518 0:default add to recovery set 1
1278010518 0:default process_node_down 1
1278010518 0:default cpg del node 1 total 2 - down
1278010518 0:default make_event_id 100020003 nodeid 1 memb_count 2 type 3
1278010518 0:default queue recover event for nodeid 1
1278010518 0:default confchg left 1 joined 0 total 2
1278010518 0:default confchg removed node 1 reason 5
Comment 12 Douglas Silas 2011-01-11 23:14:37 UTC 
    Technical note added. If any revisions are required, please edit the "Technical Notes" field
    accordingly. All revisions will be proofread by the Engineering Content Services team.
    
    New Contents:
Previously, the Closed Process Group (CPG) interface returned the wrong result, which could have led to incorrect behavior in some situations. With this update, the CPG interface now behaves as expected.
Comment 14 errata-xmlrpc 2011-01-13 23:57:02 UTC 
An advisory has been issued which should help the problem
described in this bug report. This report is therefore being
closed with a resolution of ERRATA. For more information
on therefore solution and/or where to find the updated files,
please follow the link below. You may reopen this bug report
if the solution does not work for you.

http://rhn.redhat.com/errata/RHBA-2011-0100.html