Conversation

Notices

1.  (mint@ryona.agency)'s status on Tuesday, 02-Apr-2024 01:42:16 JST 
   Back to its old tricks.
   Снимок экрана_20240401_194140.p…
   •  (mint@ryona.agency)'s status on Tuesday, 02-Apr-2024 01:43:08 JST 

     in reply to
     I'll just change the CPU governor to powersave so that the fan hum would stop driving me insane. Then it could run at 100% CPU usage however long it wants.
   •  (mint@ryona.agency)'s status on Tuesday, 02-Apr-2024 03:20:27 JST 

     in reply to

     • munir
     @munir No point, they work fine. It's a software issue.
   • munir (munir@fedi.munir.tokyo)'s status on Tuesday, 02-Apr-2024 03:20:28 JST munir

     in reply to
     @mint Oil the fans maybe
   •  (mint@ryona.agency)'s status on Tuesday, 02-Apr-2024 03:26:06 JST 

     in reply to

     • munir
     @munir My instance has been up for almost two years and is running on an underpowered hardware (although I've done a bit of work to speed some things up). That said, the issues started cropping up only last week.
   • munir (munir@fedi.munir.tokyo)'s status on Tuesday, 02-Apr-2024 03:26:08 JST munir

     in reply to
     @mint I dunno why pleroma is shitting itself on your server constantly, on mine i barely have 1% cpu usage, but then again this is single user instance
   •  (mint@ryona.agency)'s status on Tuesday, 02-Apr-2024 03:31:37 JST 

     in reply to

     • munir
     @munir Mine is proxied to a VPS via Wireguard.
   • munir (munir@fedi.munir.tokyo)'s status on Tuesday, 02-Apr-2024 03:31:38 JST munir

     in reply to
     @mint I do wonder how people manage to run services and shit from their home networks, Leased Internet Lines are so damn expensive
   • teknomunk (teknomunk@apogee.polaris-1.work)'s status on Tuesday, 02-Apr-2024 03:35:57 JST teknomunk

     in reply to

     • munir
     @mint @munir
     
     I do the same, except with tinc.
      likes this.
   • 7666 (7666@comp.lain.la)'s status on Tuesday, 02-Apr-2024 03:40:51 JST 7666

     in reply to

     • munir
     @munir @mint wazzap
      and † top dog :pedomustdie: like this.
   •  (mint@ryona.agency)'s status on Tuesday, 02-Apr-2024 04:41:26 JST 

     in reply to
     Decreased max_worker_processes to 3, CPU is still hammered a bit, but at least there's been no 500s in the last 15 minutes or so. Also it looks like the scrobble endpoint has been timing out quite often, so I nooped it in nginx for now. The long 15 seconds queries as seen in phoenix dashboard are likely timelines; maybe I should modify the frontend to hammer them less often, and add some rate limiting.
   •  (mint@ryona.agency)'s status on Tuesday, 02-Apr-2024 05:01:30 JST 

     in reply to
     Very tempted to add some RLIMIT_CPU limit to postgres container.
   •  (mint@ryona.agency)'s status on Tuesday, 02-Apr-2024 05:15:14 JST 

     in reply to
     I'll reboot the whole system in a minute, just in case.
   •  (mint@ryona.agency)'s status on Tuesday, 02-Apr-2024 05:29:35 JST 

     in reply to
     Now Beam is also shitting itself.