Re: make pg_ctl more friendly

[Junwang Zhao]

On Thu, Nov 9, 2023 at 9:57 AM Crisp Lee <litianxian...@gmail.com> wrote:
>
> Hi,
>
> I know it. But my question is not that. I did a PITR operation with 
> recovery_target_name and recovery_target_action('shutdown'). The PITR process 
> was very short and the PITR was done before pg_ctl check. The postmaster 
> shutdown due to recovery_target_action, and there was no crash. But pg_ctl 
> told me about startup failure.  I think the startup had succeeded and the 
> result was not a exception. pg_ctl should tell users about detailed messages.
>
> On Thu, Nov 9, 2023 at 9:32 AM Andres Freund <and...@anarazel.de> wrote:
>>
>> Hi,
>>
>> On 2023-11-09 09:29:32 +0800, Crisp Lee wrote:
>> > How to judge from 'DB_SHUTDOWNED' that PITR ends normally? 'DB_SHUTDOWNED'
>> > is just a state, it could not give more meaning, so I reuse the
>> > recovery.done.
>>
>> DB_SHUTDOWNED cannot be encountered while recovery is ongoing. If there was a
>> hard crash, you'd see DB_IN_ARCHIVE_RECOVERY or such, if the command was shut
>> down orderly before PITR has finished, you'd see DB_SHUTDOWNED_IN_RECOVERY.
>>
>> - Andres

After a PITR shutdown, the db state should be *shut down in recovery*, try the
patch attached.


-- 
Regards
Junwang Zhao

0001-PITR-shutdown-should-not-report-error-by-pg_ctl.patch
Description: Binary data