Retry inbox jobs for a long time
This comes from a discussion with Laura; currently the max retry amount of a job is 4, in mastodon it's 16, and usually jobs will be bounced around for a week or so, before it fully dies.
I prefer this time frame, since it makes fedi properly resilient; having AP-related jobs retry for at minimum 24 hours could account for configuration errors and server downtime. A maximum/preferred amount of time would be 1 week, since most common things will be helped in that time.
As such;
- I'd like the default to be raised to 16 or so, and/or;
- the max retry count of jobs is configurable
Summary changed:
Expand ›
State: Untriaged → Triaged
Type: Bug → Feature
Component: Queue
Component: Queue → Federation / AP Core
Component: Federation / AP Core → Queue
Component: Queue → Federation / AP Core
Target version: v2024.1-beta4
State: Triaged → Implemented
Released in version: v2024.1-beta4
Released in version: v2024.1-beta4
Released in version: v2024.1-beta4
Project
Iceshrimp.NETPriority
NormalN
Type
FeatureF
State
ImplementedI
Assignee
Laura HausmannSubsystem
BackendB
Component
Federation / AP CoreF
Target version
v2024.1-beta4V
Released in version
v2024.1-beta4V