Conversation

Notices

1. Embed this notice
   :apa: スプリットショックウイルス † (splitshockvirus@mstdn.starnix.network)'s status on Monday, 12-Aug-2024 02:11:20 JST :apa: スプリットショックウイルス †

   • かき@GNUsocialJP

   It worked, perhaps I will file a bug report for :pleromatan2:

   I'm not certain why those dependancies were needed but they weren't in the mix.exs perhaps they can be added.

   For reference, if you are using mailgun for your transactional email with pleroma you'll need plug and multipart deps added to your mix.exs

   You add them to this portion of the file, I just picked a random version of each dependancy

   defp deps do
   [
   {:plug, "~> 1.12"}, # Plug library for working with connections and query strings
   {:multipart, "~> 0.4.0"} # Multipart library for handling multipart form data
   ]
   end
   
   • ✙ dcc :pedomustdie: :phear_slackware: likes this.
   • Embed this notice
      (mint@ryona.agency)'s status on Monday, 12-Aug-2024 02:11:18 JST 

     in reply to

     • SSV
     @splitshockvirus @splitshockvirus Oban bug that freezes federation until restart: :niggachuffin:
   • Embed this notice
     :apa: スプリットショックウイルス † (splitshockvirus@mstdn.starnix.network)'s status on Monday, 12-Aug-2024 02:11:19 JST :apa: スプリットショックウイルス †

     in reply to

     • かき@GNUsocialJP
     • SSV

     @splitshockvirus@pl.starnix.network

     Memory leak is no longer present 🤞

   • Embed this notice
     SSV (splitshockvirus@pl.starnix.network)'s status on Monday, 12-Aug-2024 02:11:20 JST SSV

     in reply to
     @splitshockvirus
   • Embed this notice
      (mint@ryona.agency)'s status on Monday, 12-Aug-2024 02:19:23 JST 

     in reply to

     • SSV
     @splitshockvirus @splitshockvirus Never saw OOMs here; excluding postgres's cache memory, the system with three instances on it sits at around 2 gigs out of 24.
   • Embed this notice
     SSV (splitshockvirus@pl.starnix.network)'s status on Monday, 12-Aug-2024 02:19:24 JST SSV

     in reply to

     • 
     @mint @splitshockvirus It's ooming my VM.
   • Embed this notice
     :apa: スプリットショックウイルス † (splitshockvirus@mstdn.starnix.network)'s status on Monday, 12-Aug-2024 05:15:51 JST :apa: スプリットショックウイルス †

     in reply to

     • 
     • かき@GNUsocialJP
     • SSV

     @mint @splitshockvirus@pl.starnix.network

     Mine idle at around 500-700MBs, but as you can see it's exploded past 8GBs and into swap

     root@shikiv1:~# free -h
     total used free shared buff/cache available
     Mem: 7.6Gi 6.9Gi 146Mi 149Mi 587Mi 327Mi
     Swap: 15Gi 10Gi 5.4Gi
     root@shikiv1:~#
     

     It's probably going to crash again.

      likes this.
   • Embed this notice
     :apa: スプリットショックウイルス † (splitshockvirus@mstdn.starnix.network)'s status on Monday, 12-Aug-2024 07:29:18 JST :apa: スプリットショックウイルス †

     in reply to

     • 
     • かき@GNUsocialJP
     • SSV

     @mint @splitshockvirus@pl.starnix.network

     Applied these settings we'll see if it crashes again.

     https://docs-develop.pleroma.social/backend/configuration/optimizing_beam/

      likes this.
   • Embed this notice
      (mint@ryona.agency)'s status on Monday, 12-Aug-2024 07:31:53 JST 

     in reply to

     • SSV
     @splitshockvirus @splitshockvirus Just wanted to link this. That mostly applies to CPU and I/O busy waits, though. You can also try fiddling around with maximum heap size, though it would probably just crash the server without going OOM.
     https://www.erlang.org/doc/apps/erts/erl_cmd.html#emulator-flags