Re: Proposal: Support Logical replication of large objects

On Tue, Jan 6, 2026 at 3:05 PM Nitin Motiani <nitinmotiani(at)google(dot)com> wrote:
>
> Hi Dilip,
>
> Thanks for initiating this work. I have created a POC for the apply
> worker on top of your patch. I refactored lo_put to provide a function
> which can be called without an fmgr and used that to apply the
> lo_write operation coming from the publisher. I tested this manually
> along with your patch and it works as long as lo_create is called
> beforehand on the subscriber with the same oid.

Thanks Nitin for working on this, I will look into it.

This patch currently
> doesn't handle the tablesync worker so it can only apply the lo_write
> ops done after the replication slot is created. I'm looking into how
> to support large objects in tablesync. PFA the patch and let me know
> what you think.

Make sense, Thanks.

--
Regards,
Dilip Kumar
Google