Re: [GENERAL] Is there possibility btree_redo with XLOG_BTREE_DELETEdone between standby_redo and the end of backup

On Wed, May 24, 2017 at 11:47 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Wup, sorry, I quoted the wrong thread mentioning
> btree_xlog_delete_get_latestRemovedXid.  The right one is
>
> https://www.postgresql.org/message-id/flat/00F0B2CEF6D0CEF8A90119D4%40eje.credativ.lan
>
> and the patch is
>
> Author: Tom Lane <tgl@sss.pgh.pa.us>
> Branch: master [0e0f43d6f] 2016-08-31 08:52:13 -0400
> Branch: REL9_6_STABLE Release: REL9_6_0 [3fc489cb3] 2016-08-31 08:52:13 -0400
> Branch: REL9_5_STABLE Release: REL9_5_5 [c40bb1155] 2016-08-31 08:52:13 -0400
> Branch: REL9_4_STABLE Release: REL9_4_10 [f4e40537e] 2016-08-31 08:52:13 -0400
> Branch: REL9_3_STABLE Release: REL9_3_15 [baf111d31] 2016-08-31 08:52:13 -0400
> Branch: REL9_2_STABLE Release: REL9_2_19 [823df401d] 2016-08-31 08:52:13 -0400
> Branch: REL9_1_STABLE Release: REL9_1_24 [e3439a455] 2016-08-31 08:52:13 -0400
>
>     Prevent starting a standalone backend with standby_mode on.

Ahh. This one... Thanks. That may be related, but no way to say if
that's related with this report as there is no information that a
standalone backend is used. Perhaps that's the case though. So it
could be possible as well that the set of custom patches mentioned is
fishy.
--
Michael