[pgpool-hackers: 1271] Re: pgpool-II should fail to start when add request is rejected

Yugo Nagata nagata at sraoss.co.jp
Mon Dec 28 15:03:56 JST 2015


Hi Usama,

Although there is no comment from you, I believe the following commit
is the response to fix this.
http://git.postgresql.org/gitweb/?p=pgpool2.git;a=commitdiff;h=7f752d7606f758c5959dc91bb96346bef11de6e4;hp=8038107d8e1b0d7eb665255c11384c1a6386b94a

However, unfortunately this problem is not resolved, since SIGCHLD handeler is
not set before watchdog process is forked and pgpool waits watchdeog firing 
SIGUSR2 at sigsuspend forever.

Please find the attached patch to fix this. This set SIGCHLD handler before
initialize_watchdog() and if this fails with FATAL and exits, reaper() will
handle this and pgpool will be shutdown successfully.

Regards,

On Mon, 21 Dec 2015 19:22:44 +0900
Yugo Nagata <nagata at sraoss.co.jp> wrote:

> Usama,
> 
> When watchdog ADD request is rejected due to configuration inconsistency,
> the FATAL error occurs at the rejected watchdog. 
> 
>  FATAL:  Add to watchdog cluster request is rejected by node "localhost:9000"
> 
> However, after that, the pgpool-II parent process remains but any children
> are not forked. I can't terminate pgpool by `pgpool stop` but we have to 
> use kill command.
> 
> I think pgpool-II should fail to start and be terminated cleanly in 
> this situation. 
> 
> -- 
> Yugo Nagata <nagata at sraoss.co.jp>
> _______________________________________________
> pgpool-hackers mailing list
> pgpool-hackers at pgpool.net
> http://www.pgpool.net/mailman/listinfo/pgpool-hackers


-- 
Yugo Nagata <nagata at sraoss.co.jp>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: pgpool_main.patch
Type: text/x-diff
Size: 8757 bytes
Desc: not available
URL: <http://www.sraoss.jp/pipermail/pgpool-hackers/attachments/20151228/6b82ea24/attachment.bin>


More information about the pgpool-hackers mailing list