Open theoribeiro opened 10 months ago
I face the same issue on the production environment on GKE without autopilot the autoscaling did a scale-down but it did not come back online and had same logs errors you shared, i tried to reproduce it on a controlled environment but no luck.
I managed to resolve this by changing the limits and requests from default values in Helm chart. I increased the CPU limit from 500m to 1. Either way should be handled better.
Sorry @Technoboy- I just noticed this was assigned to me but I'm afraid I won't be able to work on it. Unassigning it :(
Search before asking
Version
Helm: 3.0 GKE: 1.27 Pulsar: 2.10.2
Minimal reproduce step
Installing a pulsar cluster on GKE autopilot using the helm chart with mostly default settings.
What did you expect to see?
A working cluster
What did you see instead?
I'm getting an exception on my brokers initialization:
Anything else?
If I disable functions using the Helm chart values, the brokers initializer correctly.
Are you willing to submit a PR?