| From: | "Higuchi, Daisuke" <higuchi(dot)daisuke(at)jp(dot)fujitsu(dot)com> |
|---|---|
| To: | 'Ramanarayana' <raam(dot)soft(at)gmail(dot)com> |
| Cc: | Laurenz Albe <laurenz(dot)albe(at)cybertec(dot)at>, "pgsql-hackers(at)postgresql(dot)org" <pgsql-hackers(at)postgresql(dot)org>, Michael Paquier <michael(at)paquier(dot)xyz>, "Tsunakawa, Takayuki" <tsunakawa(dot)takay(at)jp(dot)fujitsu(dot)com> |
| Subject: | RE: Problem during Windows service start |
| Date: | 2019-04-11 09:03:42 |
| Message-ID: | 1803D792815FC24D871C00D17AE95905DF9ADE@g01jpexmbkw24 |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Hi,
+ case POSTMASTER_STILL_STARTING:
+ write_eventlog(EVENTLOG_ERROR_TYPE, _("Timed out waiting for server startup\n"));
+ pgwin32_SetServiceStatus(SERVICE_START_PENDING);
+ return;
Could this patch solve first post's problem [1] ?
I think no one could change the service status to SERVICE_RUNNING even if the server has been started properly after timeout is occurred.
In 9.6 or earlier, the main use case where the problem is occurred is when timeout is occured because of long time recovery. Even if recovery takes a lot of time and timeout is ocurred, recovery continues in background. In this case, I want to set the status to SERVICE_RUNNING after recovery is completed.
In 10 or later, I understand wait_for_postmaster does not wait until recovery is completed, so I think this problem rarely occurs in PG 10 or later.
[1] https://www.postgresql.org/message-id/99C4246E73ED1B478BBB9671718426203E37F485@G01JPEXMBKW03
Regards,
Daisuke, Higuchi
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Chris Travers | 2019-04-11 09:25:29 | Re: Berserk Autovacuum (let's save next Mandrill) |
| Previous Message | Konstantin Knizhnik | 2019-04-11 08:46:09 | Re: Zedstore - compressed in-core columnar storage |