2 months ago
When scaling replicas up or down, ALL replicas are restarted no matter what. This causes obvious issues where reconnects are happening unnecessarily.
5 Replies
2 months ago
This thread has been opened as a bounty so the community can help solve it.
Status changed to Open Railway • about 2 months ago
0x5b62656e5d
Changing service configurations require a full redeployment.
2 months ago
It seems like most service configurations should trigger full redeployment except replica sizing.
0x5b62656e5d
Is there a reason you would not want replicas to not restart?
2 months ago
Because current long running connections like websocket/tcp connections have to re-connect causing a thundering herd.
2 months ago
brand new deployment id’s are assigned to replicas when scaling up or down. That’s how I know all replicas are being purged.