We are using AWS RDS Postgres. We reached the 200 million limit on one of the DB and triggered VACUUM FREEZE which took down the instance for roughly 40mins however the XID didn't shrink at all. We restored a snapshot of production to investigate and we were able to freeze our DBs and their age shrink significantly however the rdsadmin DB in the cluster is still at 190 million age.
Based on my understanding this freeze limit of 200million is instance wide meaning any DB reaching this limit will lead to VACUUM FREEZE.
Is rdsadmin age around 190 million a problem? Is yes, how to shrink it?
We have aggressive auto vacuum settings and we tuned as per our requirements. Our DB is vacuuming fine and its just rdsadmin by AWS which we have no control over is still causing vacuuming freeze.
We are going to reach this 200 million limit sometime this month again. Any help will be appreciated.