Mailinglist Archive: heroes (10 mails)

< Previous Next >
[heroes] New PostgreSQL cluster is now in production
Hello,

good news everyone! A long time task has finally been solved: We finally fully
migrated to a new postgresql cluster running postgresql 11.1 and leap 15.0 (the
old one was running postgresql 9 on sle11).

We get many performance, security and bug fixes, the most important being a bug
in autovacuum which resulted in consuming over 200GB of wasted disk space in
each node.

Another big improvement is the usage of a new replication manager software
named https://repmgr.org/ which allowed us to make the failover procedure just
one command!

Kudos also to Marcus Rueckert and Lars Vogdt for their valuable initial work on
repmgr and on solving various performance issues on the first migration tests.

Theo,
on behalf of the openSUSE Heroes team
< Previous Next >
This Thread
  • No further messages