Re: ERROR: subtransaction logged without previous top-level txn record

On Fri, Feb 7, 2020 at 4:29 PM Amit Kapila <amit(dot)kapila16(at)gmail(dot)com> wrote:
>
> On Wed, Feb 5, 2020 at 2:34 PM Arseny Sher <a(dot)sher(at)postgrespro(dot)ru> wrote:
> >
>
> > > What I am trying to get at is if the problem can only occur by using
> > > serialized snapshots and we fix it by somehow not using them initial
> > > slot creation, then ideally we don't need to remove the error in
> > > ReorderBufferAssignChild, but if that is not the case, then we need to
> > > discuss other cases as well.
> >
> > So, 1) and 3) mean this is not the case.
> >
>
> Right, I am thinking that if we can find some way to skip the xact
> assignment for (1) and (3), then that might be a better fix.
>

Just to be clear, I am just brainstorming the ideas to see if we can
find some better solutions to the problems (1) and (3) described by
Arseny in the above email [1]. At this stage, it is not clear to me
that we have a fix simple enough to backpatch apart from what Arseny
posted in his fist email [2] (which is to stop demanding that top xact
must be seen before subxact in decoding.).

[1] - https://www.postgresql.org/message-id/87zhdx76d5.fsf%40ars-thinkpad
[2] - https://www.postgresql.org/message-id/87ftjifoql.fsf%40ars-thinkpad

--
With Regards,
Amit Kapila.
EnterpriseDB: http://www.enterprisedb.com