RE: Slow catchup of 2PC (twophase) transactions on replica in LR - Mailing list pgsql-hackers

From Zhijie Hou (Fujitsu)
Subject RE: Slow catchup of 2PC (twophase) transactions on replica in LR
Date
Msg-id OS3PR01MB57184E0995521300AC06CB4B94A72@OS3PR01MB5718.jpnprd01.prod.outlook.com
Whole thread Raw
In response to RE: Slow catchup of 2PC (twophase) transactions on replica in LR  ("Hayato Kuroda (Fujitsu)" <kuroda.hayato@fujitsu.com>)
List pgsql-hackers
On Tuesday, July 9, 2024 8:53 PM Hayato Kuroda (Fujitsu) <kuroda.hayato@fujitsu.com> wrote:
> 
> > 0001 - Codes for SUBOPT_TWOPHASE_COMMIT are moved per requirement
> [1].
> >        Also, checks for failover and two_phase are unified into one function.
> > 0002 - updated accordingly. An argument for the check function is added.
> > 0003 - this contains documentation changes required in [2].
> 
> Previous patch set could not be accepted due to the initialization miss.
> PSA new version.

Thanks for the patches ! I initially reviewed the 0001 and found that
the implementation of ALTER_REPLICATION_SLOT has a issue, e.g.
it doesn't handle the case when there is only one specified option
in the replication command:

ALTER_REPLICATION_SLOT slot (two_phase)

In this case, it always overwrites the un-specified option(failover) to false even
when the failover was set to true. I tried to make a small fix which is on
top of 0001 (please see the attachment).

I also added the doc of the new two_phase option of the replication command
and a missing period of errhint in the topup patch.

Best Regards,
Hou zj

Attachment

pgsql-hackers by date:

Previous
From: Alexander Cheshev
Date:
Subject: Re: [PATCH] TODO “Allow LISTEN on patterns”
Next
From: Pavel Luzanov
Date:
Subject: Re: Things I don't like about \du's "Attributes" column