View Issue Details

IDProjectCategoryView StatusLast Update
0000537Pgpool-IIBugpublic2019-08-26 09:49
Reportergmartinez Assigned Topengbo  
PrioritynormalSeveritymajorReproducibilityalways
Status closedResolutionopen 
PlatformLinuxOSCentOSOS Version7
Product Version3.7.10 
Summary0000537: pgpool no longer executing if_down_cmd on shutdown
DescriptionWe migrated from CentOS 6 + pgpool 3.7.7 to CentOS 7 and pgpool 3.7.10, and after this upgrade it seems that pgpool is no longer executing if_down_cmd on shutdown/restart.
We see it still executing if_up_cmd when a master is elected, but not if_down_cmd on the previous master.

Steps To Reproduce1) Start pgpool cluster
2) Run `systemctl stop pgpool.service` on the watchdog master
3) Notice that if_down_cmd did not execute on the old master.
Additional InformationWhen checking logs for anything mentioning if_down_cmd after the upgrade), we only see this:

WARNING: checking setuid bit of if_down_cmd

(which is OK, this command doesn't need setuid)

We do see the corresponding up command being executed in the logs:

Aug 09 10:22:34 db006 cgexec[597088]: 2019-08-09 10:22:34: pid 599161: DETAIL: 'if_up_cmd' returned with success


I've attached the systemd unit file + the sysconfig file.
TagsNo tags attached.

Activities

gmartinez

2019-08-09 20:37

reporter  

sysconfig (175 bytes)   
sysconfig (175 bytes)   
pgpool.service (440 bytes)

pengbo

2019-08-16 10:50

developer   ~0002788

Can you see the log messages below in pgpool log?

---------
watchdog: de-escalation started
...
watchdog de-escalation failed to bring down delegate IP
---------


Can you execute "pgpool_wd_state up" and "pgpool_wd_state down" successfully on CentOS 7?

gmartinez

2019-08-19 21:29

reporter   ~0002795

Hmm, curious. It looks like it worked after the next restart...


Aug 12 11:19:52 db006 cgexec[196062]: 2019-08-12 11:19:52: pid 1380400: LOG: watchdog: de-escalation started
Aug 12 11:19:52 db006 cgexec[196062]: 2019-08-12 11:19:52: pid 196063: LOG: watchdog de-escalation process with pid: 1380400 exit with SUCCESS.
Aug 12 11:19:52 db006 cgexec[196062]: 2019-08-12 11:19:52: pid 1380400: DETAIL: 'if_down_cmd' returned with success

pengbo

2019-08-22 15:47

developer   ~0002797

If you have resolved this problem, may I close this issue?

gmartinez

2019-08-22 22:41

reporter   ~0002798

Yes, please close it.

Issue History

Date Modified Username Field Change
2019-08-09 20:37 gmartinez New Issue
2019-08-09 20:37 gmartinez File Added: sysconfig
2019-08-09 20:37 gmartinez File Added: pgpool.service
2019-08-14 11:17 pengbo Assigned To => pengbo
2019-08-14 11:17 pengbo Status new => assigned
2019-08-14 11:17 pengbo Description Updated
2019-08-14 11:17 pengbo Steps to Reproduce Updated
2019-08-14 11:17 pengbo Additional Information Updated
2019-08-16 10:50 pengbo Note Added: 0002788
2019-08-16 10:50 pengbo Status assigned => feedback
2019-08-19 21:29 gmartinez Note Added: 0002795
2019-08-19 21:29 gmartinez Status feedback => assigned
2019-08-22 15:47 pengbo Note Added: 0002797
2019-08-22 22:41 gmartinez Note Added: 0002798
2019-08-26 09:49 pengbo Status assigned => closed