Summary
After upgrading Airbyte to version 1.1.0 via Helm in Kubernetes, syncs are not running successfully. Replication-job pods are created but remain in Pending status due to ‘Insufficient cpu’ error.
Question
Hello, after upgrading to 1.1.0 (via Helm in K8s) Airbyte is not able to run syncs, the replication-job pods gets created and start hanging in Pending status, the error I see is 0/5 nodes are available: 5 Insufficient cpu. preemption: 0/5 nodes are available: 5 No preemption victims found for incoming pod.
This topic has been created from a Slack thread to give it more visibility.
It will be on Read-Only mode here. Click here if you want
to access the original thread.
Join the conversation on Slack
["airbyte", "upgrade", "1.1.0", "kubernetes", "sync-issue", "replication-job", "pending-status", "insufficient-cpu"]
Which version did you have before?
Previously, read & write, they were separate pods, but since 1.0 (I think) there are more containers in replication-job (init, orchestrator, source, destination). I guess you need more CPU for your nodes, to be able to schedule those replication-job pods
yeah I noticed there are more containers indeed, so I need to add resource request to the workload-launcher, or the worker?
How much cpu in Capacity/Allocatable do you get for your nodes kubectl describe nodes?
e.g.
cpu: 8
...
Allocatable:
cpu: 8
...```
cpu: 4
...
Allocatable:
cpu: 3920m
...```
and it seems it needs at least five…
Like in error, Insufficient cpu
Thanks for your help, didn’t think about the cluster resources, I’m not directly managing that
you can also try to modify requests/limits for jobs
jobs:
resources:
requests:
cpu: ...
memory: ...
limits:
cpu: ...
memory: ...```
yeah, I’ll try that first
do you have any suggestion for that?
empirically 
I would recommend installing k9s https://k9scli.io/topics/install/ to check current usage and you can adjust accordingly