Parametric analysis does not start

asked 2023-01-18 02:19:54 -0600

MapMortar
145 ●2 ●6 https://www.mapmortar.io/

updated 2023-01-20 03:51:16 -0600

I am struggling to test PAT 3.5.1.

I have deployed the cluster on AWS and verified that all the pods are running

NAME                                                       READY   STATUS    RESTARTS   AGE
db-65f4c9ccf6-bzj4f                                        1/1     Running   0          30m
openstudio-server-nfs-server-provisioner-7784b767d-zdldd   1/1     Running   0          30m
redis-5ffc4f5d9f-lpcg6                                     1/1     Running   0          30m
rserve-5764d6f988-mqkvq                                    1/1     Running   0          30m
web-7d758b7fc5-trpwx                                       1/1     Running   0          30m
web-background-764bbfbc77-tfjc9                            1/1     Running   0          30m
worker-59cc4694b5-5bs2h                                    1/1     Running   0          30m
worker-59cc4694b5-9gbbg                                    1/1     Running   0          29m

I can connect to the server and view it on a browser, but no project is shown. If I show the development tools I can see the following error message when I launch the workflow. The analysis seems to start, but it gets stuck at Creating Analysis zip.

image description

It is a simple algorithmic analysis (LHS) with only one measure.

I have different AWS accounts and the cluster is created with the --profile user option. Could this be an issue? Does PAT uses the default AWS credentials in .aws?

Any help would be appreciated.

EDIT 1 Not sure if this is relevant, but when I check the cluster on AWS I see this message.

Your current user or role does not have access to Kubernetes objects on this EKS cluster This may be due to the current user or role not having Kubernetes RBAC permissions to describe cluster resources or not having an entry in the cluster’s auth config map.Learn more

Also, the ccontent of ~/.kube/config is the following

apiVersion: v1
clusters:
- cluster:
    certificate-authority-data: ......
    server: https://C63FA7D546267F20444E92C67EF9BABA.yl4.eu-west-2.eks.amazonaws.com
  name: openstudio-server.eu-west-2.eksctl.io
contexts:
- context:
    cluster: openstudio-server.eu-west-2.eksctl.io
    user: ruggiero@openstudio-server.eu-west-2.eksctl.io
  name: ruggiero@openstudio-server.eu-west-2.eksctl.io
current-context: ruggiero@openstudio-server.eu-west-2.eksctl.io
kind: Config
preferences: {}
users:
- name: ruggiero@openstudio-server.eu-west-2.eksctl.io
  user:
    exec:
      apiVersion: client.authentication.k8s.io/v1beta1
      args:
      - token
      - -i
      - openstudio-server
      command: aws-iam-authenticator
      env:
      - name: AWS_STS_REGIONAL_ENDPOINTS
        value: regional
      - name: AWS_DEFAULT_REGION
        value: eu-west-2
      - name: AWS_PROFILE
        value: mioruggierodev
      provideClusterInfo: false

EDIT 2 The server seems to be working. I have tested it with the urbanoopt optimization work flow using the provided test files.

A project is created and the simulations run. They all fail, but that is a problem for another day

EDIT 3 (Adding here for convenience.) More tests. I opened the same file in PAT 3.4.0. The OSA and zip files are now generated without any issue. However, the workflow fails with this error

image description

But if I use openstudio_meta to run the analysis with

/Applications/ParametricAnalysisTool-3.5.1/ParametricAnalysisTool.app/Contents/Resources/ruby/bin/ruby /Applications/ParametricAnalysisTool-3.5.1/ParametricAnalysisTool.app/Contents/Resources/OpenStudio-server/bin/openstudio_meta run_analysis --debug --verbose testpat01.json http://aa98744be7a3c47fcadb63a78e8e493f-1360380086.eu-west-2.elb.amazonaws.com -a lhs

Everything works as expected and the analysis completes on the server.

edit retag flag offensive close merge delete

add a comment

answered 2023-01-18 10:10:31 -0600

tijcolem
161 ●1 ●2

Hi Aaron,

I just want to confirm a couple of things before offering some options and/or filing a bug report. From your description, it sounds like you are having a problem submitting the analysis from PAT to a cloud instance of OpenStudio-server? So I just want to confirm that you are able to connect to the running OpenStudio cloud instance by inputting the url address and that the server status icon in PAT is green? See example screenshot below.

C:\fakepath\Screen Shot 2023-01-18 at 9.08.25 AM.png

edit flag offensive delete link

Comments

Hi, I have deleted the cluster now, but yes, I can confirm that the the icon was green. Also I can see the server from both the PAT tab and from a browser using the link generated by AWS, but there is no project.