On Thu, May 7, 2020 at 1:33 PM Michael Lewis <mlewis@xxxxxxxxxxx> wrote:
It is trying to do a vacuum freeze. Do you have autovacuum turned off? Any settings changed from default related to autovacuum?Read 24.1.5. Preventing Transaction ID Wraparound FailuresThese may also be of help-
https://info.crunchydata.com/blog/managing-transaction-id-wraparound-in-postgresql
https://www.2ndquadrant.com/en/blog/managing-freezing/Note that you need to ensure the server gets caught up, or you risk being locked out to prevent data corruption.
Thanks Mike.
1) We haven't changed anything related to autovacuum except a work_mem parameter which was increased to 4 GB which I believe is not related to autovacuum
2) The vacuum was not turned off and few parameters we had on vacuum are
autovacuum_analyze_scale_factor = 0.02 and autovacuum_vacuum_scale_factor = 0.05
3) The database curently we are running is 2 years old for now and we have around close to 40 partitions and the datfrozenxid on the table is 343 million whereas the default is 200 million. I would try doing a manual auto vacuum on those tables
where the autovacuum_freeze_max_age > 200 million. Do you think It's a right thing to do ?.
I will also go through this documents.
Tahnks
