[pgpool-general: 7689] Re: Segmentation fault in watchdog child process

[Emond Papegaaij]

Hi,

I've ran the full test suite 3 times with 4.2.3 and so far I haven't seen
any segmentation faults on any of the nodes. All runs with 4.2.4 I've
checked the logs for, have at least one segmentation fault on one of the
nodes, often several. So it seems this segfault is a regression somewhere
between 4.2.3 and 4.2.4.

Best regards,
Emond

On Wed, Sep 8, 2021 at 3:37 PM Emond Papegaaij <emond.papegaaij at gmail.com>
wrote:

> Hi,
>
> While running our integration tests, we are seeing frequent failures in
> the failovers. Inspecting the Pgpool logs revealed that these failures seem
> to caused by the leader watchdog child process segfaulting in the middle of
> the failover. This often results in a partial failover and an inconsistent
> database cluster. I've attached a dump of the Ppgool (version 4.2.4) docker
> container logs where you can see 5 instances of a segmentation fault in the
> watchdog process.
>
> Overnight I'll run the same tests against 4.2.3 to see if this is a
> regression in 4.2.4. Please let me know if you need additional information
> (and how to obtain that) to further troubleshoot this issue.
>
> Best regards,
> Emond
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.pgpool.net/pipermail/pgpool-general/attachments/20210908/289b3333/attachment.htm>