Hi all. We are running dagster using kube containe...
# ask-community
d
Hi all. We are running dagster using kube containers. And we get this error message in the user deployment pod: Readiness probe errored: command “dagster api grpc-health-check -p 3030” timed out Because of this, our production is down. Anyone can help?
🤖 1
m
Does it happen for whole deployment or the single pod (runner)? As far I understand with Docker Runner (container per repository), you need to ensure that container that is used for repository has no errors (eg. missing files or entry points).
d
it’s only for the single pod
Is it because dagit pod isn’t aware of the new deployed user deployment pod?
several hours ago, I tried deploying the “exact same” user deployment code into staging. But it’s just not working in production.
m
Does the code from this pod, works with dagit locally or not? 🙂
I mean, you should be able to use the repository aka user code locally with dagit without having to deploy it straight into K8s or Docker Swarm
If it works, maybe you should consider sharing the deployment setup with K8s runner to ensure that it is properly set-up
d
The “same code” has been deployed to staging as well. And totally works.
Sorry. What exactly does this message mean? What is it supposed to do? ” Readiness probe errored: command “dagster api grpc-health-check -p 3030" timed out ”
m
As far I understand, this means that the configuration for this deployment on production differs from this on staging and pod doesn't start. What is the output of the pod that is not working?
d
Sorry. Forgot to update you that issue was resolved a while ago. Thanks