Recommended Kubernetes Cluster Configurations/Sizing

google
openstudio-server
pat
retag add tags

Hi,

I have implemented openstudio-server using the Google Cloud framework and encountered some connection difficulties. When I followed the documentation on google and OpenStudio Server the first time, I was able to set up the OpenStudio Server and run a complete analysis. I then deleted the cluster to save costs. Afterward, I went through the same process several times but failed to get the server running properly; thus, I can't connect to the server via PAT. I have got a message, "Pod unschedulable", on the Google Cloud Console in which it seems like the cluster was not created with enough resources.

Therefore, could anyone please advise on creating a cluster with sufficient resources for a PSO analysis (e.g., disk size, machine type, and number of nodes)?

Also, is there a command to restart the "pending" pods below since the "get pods" shows the RESTARTS column?

Thank you so much.

kubectl get pods

NAME                                                        READY   STATUS    RESTARTS   AGE
db-759856dc8c-m6qw8                                         0/1     Pending   0          22m
openstudio-server-nfs-server-provisioner-56ccb8b855-hm4jf   0/1     Pending   0          22m
redis-5b8c9599d6-4jd4c                                      0/1     Pending   0          22m
rserve-6897979876-fz944                                     0/1     Pending   0          22m
web-6bd89749cf-5wkbs                                        0/1     Pending   0          22m
web-background-584868f998-dqqjw                             0/1     Pending   0          22m
worker-84fb55698b-722bs                                     1/1     Running   0          22m
worker-84fb55698b-xpt5w                                     1/1     Running   0          22m

asked 2023-07-02 05:57:19 -0500, updated 2023-07-02 07:55:22 -0500

edit flag offensive 0 remove flag close merge delete

Comments

add a comment see more comments

2 Answers

Sort by

oldest

newest

most voted

Hi,

A lot of the times pending state is due to insufficient resources or volume issues. Can you run cmd kubectl describe pod <podid> and see what it says?

161

tijcolem

answered 2023-07-10 09:13:39 -0500

edit flag offensive 0 remove flag delete link

Comments

add a comment see more comments

thus, I can't connect to the server via PAT. I have got a message, "Pod unschedulable", on the Google Cloud Console in which it seems like the cluster was not created with enough resources.

My 2 cents, is it possible there is some limit got hit from GCP?

Could you try to restart it from the scractch (i.e. delete the node and renew one from GCP) and see how it works again.

I suspect there is some resource(maybe disk) used up in your node/box/server.

Thanks,