Orchestrator pod orchestrator-repl-job non-existent. Assume failure.

[Alex Vdovyn (avdovyn)]

We've noticed following error:
"Terminal state in the State Store is missing. Orchestrator pod orchestrator-repl-job non-existent. Assume failure."
Which is quite frequent during the jobs run.

Could you advise what might be wrong in the configuration?
Or how this issue can be remediated.

A few notes about our implementation:

• The current version of Airbyte is 0.54.69
• We are running Airbyte on AWS EKS cluster (Kubernetes)
• EKS is configured to run on 2xlarge instances, with maximum number of nodes is 45. While minimum is 1.
• Specs for the jobs pods are: Request: 6Gb of memory, Limit: 8Gb
• This error happens to a long running Teams connector (it pull data from Graph API)