hello, I'm back. We've upgraded airbyte to version...
# replication-troubleshooting
e
hello, I'm back. We've upgraded airbyte to version “0.40.18" from "0.35." and now a lot of jobs get stuck for over 30 mins. Last log entry is something like.
Copy code
2022-10-20 13:18:21 INFO i.a.c.i.LineGobbler(voidCall):114 - 
2022-10-20 13:18:21 INFO i.a.c.i.LineGobbler(voidCall):114 - ----- END CHECK -----
2022-10-20 13:18:21 INFO i.a.c.i.LineGobbler(voidCall):114 -
and then nothing for over half an hour, but the sync is still running. we're using kubernetes, deploying with helm. what can we do to improve performance? increase job resource limits?
✍️ 1
u
@[DEPRECATED] Marcos Marx turned this message into Zendesk ticket 2983 to ensure timely resolution!
e
also a lot of jobs get "stuck" with the last entry "2022-10-20 141346 destination > Completed integration: io.airbyte.integrations.destinations..." for over an hour
Or at
2022-10-20 14:13:47 INFO i.a.c.t.TemporalUtils(withBackgroundHeartbeat):280 - Stopping temporal heartbeating...
h
Hey when you say the sync is still running you mean the pod is still up and is in running state?
e
hi, I meant airbyte ui showed the sync as still running
there were a lot of pods though
none of them were in the failed state
a lot of jobs did seem to have successfully replicated data before "hanging"
h
Got it. Seems to me like a resource issue. Do you mind testing with just one sync and then see how the pod status is changing
e
a small sync ran successfully from start to finish
I agree about the resources, but how do we calculate the right amount of resources? do you have recommendations?
h
Hey maybe this doc can help you there https://docs.airbyte.com/operator-guides/scaling-airbyte