Swarm.Tracker stuck in `syncing` state when lots of nodes starting/crashing
Open
- Dominant language
- Elixir
- Stars
- 1.3k
- Forks
- 110
- PR merge metrics
- No merged PRs in 30d
Description
I can't reproduce this on reliably, but I've noticed that occasionally `Swarm.Tracker` gets stuck in the `syncing` state when a lot of nodes are crashing & restarting.
Based on inspecting the process on each surviving node last time I saw it, it looks like they are stuck in the `syncing` state, and have a list of `pending_sync_requests`, including PIDs from nodes no longer in the cluster. (I think)
Contributor guide
No contributing guide indexed for this repository
Assessment
This issue has not been assessed yet.