Re: Minimal logical decoding on standbys

Hi,

On 1/11/23 4:23 PM, Drouvot, Bertrand wrote:
> Hi,
>
> On 1/11/23 8:32 AM, Bharath Rupireddy wrote:
>> On Tue, Jan 10, 2023 at 2:03 PM Drouvot, Bertrand
>> <bertranddrouvot(dot)pg(at)gmail(dot)com> wrote:
>>>
>>> Please find attached, V37 taking care of:
>>
>> Thanks. I started to digest the design specified in the commit message
>> and these patches.
>
> Thanks for looking at it!
>
>> Here are some quick comments:
>>
>> 1. Does logical decoding on standby work without any issues if the
>> standby is set for cascading replication?
>>
>
> Without "any issues" is hard to guarantee ;-) But according to my tests:
>
> Primary -> Standby1 with or without logical replication slot -> Standby2 with or without logical replication slot
>
> works as expected (and also with cascading promotion).
> We can add some TAP tests in 0004 though.

Cascading standby tests have been added in V48 attached.

It does test that:

- a sql logical decoding session on the cascading standby get expected output before/after the standby promotion
- a pg_recvlogical logical decoding session on the cascading standby (started before the standby promotion) get expected output before/after the standby promotion

Regards,

--
Bertrand Drouvot
PostgreSQL Contributors Team
RDS Open Source Databases
Amazon Web Services: https://aws.amazon.com