We’ll give the upgrade new try tomorrow. I’ve had some good input from admins of other instances, which are also gonna help troubleshoot during/after the upgrade.

Also there are newer RC versions with fixed issues.

Be aware that might we need to rollback again, posts posted between the upgrade and the rollback will be lost.

We see a huge rise in new user signups (duh… it’s July 1st) which also stresses the server. Let’s hope the improvements in 0.18.1 will also help with that.

  • @Jordynl
    link
    471 year ago

    Can’t you disable the post button within the upgrade progress? It’s a shame to see posts to disappear due to it.

    • @innermeerkat
      link
      91 year ago

      The best would be to redirect to a 502 for the duration of the migration. Better safe than sorry

      • @Mereo
        link
        3
        edit-2
        1 year ago

        I think a static page page explaining to users that the server is being upgraded would be best.

        • @innermeerkat
          link
          31 year ago

          Exactly. That’s what I meant by a 502. Like most HTTP responses, you can associate a template with it and it’s even recommended when you have a lot of people using your service and you want to explain why it’s not accessible right now.

        • BitOneZero @ .world
          link
          31 year ago

          I think a static page page explaining to users that the server is being upgraded would be best.

          Beehaw.org did this, their nginx seems configured to return a nice error message.

    • @Chocrates
      link
      51 year ago

      It’s not the upgrade that is the problem it is the rollback. And we can’t test the upgrade without posting.

    • BitOneZero @ .world
      link
      11 year ago

      Can’t you disable the post button within the upgrade progress?

      No. And I think the ability to put the site into read-only mode for upgrade shakedown, be it hardware, operating system, or Lemmy upgrade is important for Lemmy to get past these “growing pains”.