https://dagster.io/ logo
Title
a

Alejandro DE LA CRUZ LOPEZ

07/09/2021, 9:09 AM
Hello everyone! I'm having a very weird issue. When deploying dagster locally on minikube it works perfectly but when I try to deploy it with the same configuration in my company's k8 cluster it goes slowly and when I try to run pipelines it appears "starting" forever. I get this message in the port-forward command:
E0709 11:02:52.896413  378418 portforward.go:372] error copying from remote stream to local connection: readfrom tcp6 [::1]:8080->[::1]:35554: write tcp6 [::1]:8080->[::1]:35554: write: broken pipe
I attach an image of the forever "starting" and the quota of the namespace (in case it's useful). Do you have any idea where the issue could be?
d

daniel

07/12/2021, 1:15 PM
Hi Alejandro, sorry for the delay here. Is it possible to send the output of 'kubectl logs' from the pod for the job that it references in that screenshot?
the name of the pod should be the same as that kubernetes job name, but with an additional suffix at the end
a

Alejandro DE LA CRUZ LOPEZ

07/12/2021, 1:29 PM
The pod nor the job are ever created. I have been trying different stuff and by setting
loadInclusterConfig
to false on
values.yaml
seems to work. I am not getting those issues when deploying it locally on minikube though. Thanks for the reply