> On Nov 12, 2018, at 10:28 PM, Andrey Zhidenkov <andrey.zhidenkov@gmail.com> wrote:
>
> We have a few database clusters (1 master, 2 hot standbys) with
> hot_standby_feedback setting on. But I am constantly seeing query
> conflicts on standbys because dead rows on the masters are deleted by
> VACUUM. Usually I notice it shortly after VACUUM ends, because the
> replication lag on replicas is starting to increase.
>
> vacuum_defer_cleanup_age = 0. PostgreSQL version is 9.5.13. Cluster
> load is about 1500 transactions per second.
>
> Any ideas?
>
> -
> With the best regards, Andrey Zhidenkov
>
Are these streaming replicas? The wal_streaming process on the master should show the xmin of the replicas; you can
checkthe value from pg_stat_activity to make sure the feedback is occurring.
Are the replicas disconnecting from the replication stream? As the would mean the xmin from the replica would no longer
bein play.