oldsysops

@oldsysops@social.dk-libre.fr

main account of @oldsysops@mamot.fr
0 ★ 4 ↺

[?]oldsysops » 🌐
@oldsysops@social.dk-libre.fr

: when you refresh collation on a database after an os upgrade, is there a way to refresh collation on the replicate without launching a full pg_basebackup ?


    ...

    [?]Julien Riou » 🌐
    @jriou@hachyderm.io

    @oldsysops If it's a physical replica, the database collation is updated on the replica as soon as you update the primary.

      ...
      0 ★ 0 ↺

      [?]oldsysops » 🌐
      @oldsysops@social.dk-libre.fr

      @jriou@hachyderm.io yes,
      i've just figure out that i forgot to upgrade the secondary server so the replication was not working...
      my mistake

        ...

        [?]Julien Riou » 🌐
        @jriou@hachyderm.io

        @oldsysops Upgrade the OS or PG on the replica?

          ...

          [?]Julien Riou » 🌐
          @jriou@hachyderm.io

          @oldsysops OK, I thought we could run pg_upgrade on the replica for a second but it's not possible so I guess it's the OS

            1 ★ 0 ↺

            [?]oldsysops » 🌐
            @oldsysops@social.dk-libre.fr

            @jriou@hachyderm.io the os, i've done an upgrade to debian 13, switch to the upgraded secondary and the replication log was all about collation mismatch...
            but that's ok now, everything is up to date now and replication is on.
            i can rest now 🙂

              ...

              [?]Julien Riou » 🌐
              @jriou@hachyderm.io

              @oldsysops You should reindex all databases concurrently after a collation refresh

                ...
                0 ★ 0 ↺

                [?]oldsysops » 🌐
                @oldsysops@social.dk-libre.fr

                @jriou@hachyderm.io already done !
                it was on my check-list.

                  Historique