We just released a new patch which fixes an important federation problem. Due to a mistake in the code, activities were only sent out once, even if delivery failed (e.g. because the target instance is down, or a temporary connection issue). Instance admins should definitely upgrade.

This problem is fixed, so that failed deliveries are retried up to 9 times, with exponentionally increasing intervals. The only limitation now is that it will not keep retrying after Lemmy is restarted, but that case should be less common.

PR link

  • seahorse
    9·
    3 years ago

    I’m still having federation issues after the upgrade. At this point I wonder if it’s my firewall or something. I watch the lemmy logs when upvoting for example and my server doesn’t even see the request come in.

    • nutomicOPMA
      8·
      3 years ago

      Damn, it looks like there are additional problems with federation. We will release 0.16 in the next days which will include additional logging, and allow me to debug this.

      To rule out the firewall, the main thing you need to ensure is that POST requests to /inbox go through.

      • seahorse
        2·
        3 years ago

        Yeah, according to logs the POST requests are going through fine.

        To                         Action      From
--                         ------      ----
22/tcp                     ALLOW IN    Anywhere
80/tcp                     ALLOW IN    Anywhere
443                        ALLOW IN    Anywhere
25                         DENY IN     Anywhere
443/tcp                    ALLOW IN    Anywhere
22/tcp (v6)                ALLOW IN    Anywhere (v6)
80/tcp (v6)                ALLOW IN    Anywhere (v6)
443 (v6)                   ALLOW IN    Anywhere (v6)
25 (v6)                    DENY IN     Anywhere (v6)
443/tcp (v6)               ALLOW IN    Anywhere (v6)

443                        ALLOW OUT   Anywhere
443 (v6)                   ALLOW OUT   Anywhere (v6)