@Gargron > Is there any way to clean database for now offlined instances filling sidekiq logs of errors and hammering networks ? `tootctl accounts cull` seems not addressing that.

@Gargron Arround 8 errors per second at this moment… 😱
And I'm just a tiny single user instance…

Show thread

@Gargron Got my /var/log full of sidekiq error log 😭
From 17/11 12:51 to 18/11 19:07, 5.8GB of logs, 2700 retries per hosts… 😱
Now my 20 sidekiq queues are full, need to upgrade to 50…

Show thread

@aeris Stoplight handles short-term errors and is meant to function this way. These might not be offline but temporarily out of order. If you know something is offline for sure and want to delete it you can use `tootctl domains purge DOMAIN`

@Gargron So « short-term » not working 😂 is down from decades (at least july)

@Gargron Got actually 4318 errors in 10 minutes. Sidekiq full of retry, handle more of useless/broken feeds than usefull ones. Have to upgrade to 20 workers to be able to handle the traffic, 5 was no more enough and sidekiq queue growing 😭

@aeris Was your instance down for a long time or how come?

@Gargron Not down since 2 days, and last maintenance didn't last more than 1 hour.

@aeris @valere @Gargron Ca aurait été peut-être un peu plus rapide de faire tootctl purge <domaine> <domaine> <domaine> <domaine> <domaine>.... 🤔

@aeris @valere @Gargron OK alors ^_^

Perso ça m'a prit à peine une heure, mais on ne doit pas avoir le même niveau d'instance #scalabilité

Sign in to participate in the conversation

The social network of the future: No ads, no corporate surveillance, ethical design, and decentralization! Own your data with Mastodon!