Jupyter based workspace is not working

Describe the bug

Che version

Latest

Steps to reproduce

After running the command to instanciate an eclispse multi-user, single-host on kubernetes using an helm template (used to enable a storageCLassName for the keyclock PVC).

chectl server:start --platform=k8s --installer=helm --multiuser --domain=che-dev.esa-maap.org --k8spodwaittimeout=400000 --postgres-pvc-storage-class-name="sata-iac-0b" --workspace-pvc-storage-class-name="sata-iac-0b" --templates=/home/cloud/chectl-templates/DEV/templates

We have a fresh instance of che running. When we tried to create a workspace, it failed without any log in the pod up and down quickly (5s):

workspacelcdkxz0gq8s7o61d.che-plugin-broker   0/1   Pending   0     0s
workspacelcdkxz0gq8s7o61d.che-plugin-broker   0/1   Pending   0     0s
workspacelcdkxz0gq8s7o61d.che-plugin-broker   0/1   ContainerCreating   0     0s
workspacelcdkxz0gq8s7o61d.che-plugin-broker   0/1   Terminating   0     5s
workspacelcdkxz0gq8s7o61d.che-plugin-broker   0/1   Terminating   0     5s

In the ingress controller we can see :

172.16.0.49 - - [03/Sep/2020:13:07:14 +0000] "GET /api/workspace/workspaceoma0eirhe30xk4vg HTTP/2.0" 200 1466 "https://che-dev.esa-maap.org/workspace-loader/admin/angular-bln86?uid=10718" "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/84.0.4147.125 Safari/537.36" 1517 0.010 [che-che-host-8080] [] 172.16.0.16:8080 3512 0.009 200 746f46d32a2baa0f9d27b01be55143d0
W0903 13:08:06.845532       8 controller.go:919] Service "che/server2p05oun5-jwtproxy" does not have any active Endpoint.
W0903 13:08:06.845557       8 controller.go:919] Service "che/servertauyur1g-theia-ideaty" does not have any active Endpoint.
W0903 13:08:06.845569       8 controller.go:919] Service "che/servertauyur1g-theia-ideaty" does not have any active Endpoint.
W0903 13:08:06.845580       8 controller.go:919] Service "che/server2p05oun5-jwtproxy" does not have any active Endpoint.
W0903 13:08:06.845590       8 controller.go:919] Service "che/server2p05oun5-jwtproxy" does not have any active Endpoint.
W0903 13:08:06.845602       8 controller.go:919] Service "che/servertauyur1g-theia-ideaty" does not have any active Endpoint.

The k8s event seems good :

72s         Normal   SuccessfulMountVolume   Pod       Successfully mounted volumes for pod "workspaceoma0eirhe30xk4vg.che-plugin-broker_che(3c7f1ff8-edf4-11ea-801b-fa163e6c591b)"
72s         Normal   Scheduled               Pod       Successfully assigned che/workspaceoma0eirhe30xk4vg.che-plugin-broker to 192.168.0.208
71s         Normal   Pulling                 Pod       pulling image "quay.io/eclipse/che-plugin-metadata-broker:v3.3.0"
69s         Normal   Started                 Pod       Started container
69s         Normal   SuccessfulCreate        Pod       Created container
70s         Normal   Pulled                  Pod       Successfully pulled image "quay.io/eclipse/che-plugin-metadata-broker:v3.3.0"
68s         Normal   CREATE                  Ingress   Ingress che/ingressvotn16ea
68s         Normal   CREATE                  Ingress   Ingress che/ingressv4gp0r51
68s         Normal   CREATE                  Ingress   Ingress che/ingresstyrs0l3s
68s         Normal   CREATE                  Ingress   Ingress che/ingressr348sl0a
68s         Normal   CREATE                  Ingress   Ingress che/ingressljd7sr5b
68s         Normal   CREATE                  Ingress   Ingress che/ingressjrkfr8vh
68s         Normal   CREATE                  Ingress   Ingress che/ingressfdnal9k9
68s         Normal   CREATE                  Ingress   Ingress che/ingress9plc5czm

The share folder for the workspace contains en empty file for the log : /mnt/dev-iac/DEV/che/workspaceoma0eirhe30xk4vg/che-logs-che-plugin-broker

After some minutes the che log contains:

2020-09-03 13:03:02,198[nio-8080-exec-9]  [INFO ] [o.e.c.a.w.s.WorkspaceRuntimes 480]   - Starting workspace 'admin/angular-bln86' with id 'workspaceoma0eirhe30xk4vg' by user 'admin'
2020-09-03 13:08:06,786[aceSharedPool-8]  [WARN ] [.i.k.KubernetesInternalRuntime 247]  - Failed to start Kubernetes runtime of workspace workspaceoma0eirhe30xk4vg. Cause: Waiting for ingress 'ingress92p434bq' reached timeout
2020-09-03 13:08:07,047[aceSharedPool-8]  [INFO ] [o.e.c.a.w.s.WorkspaceRuntimes 984]   - Workspace 'admin:angular-bln86' with id 'workspaceoma0eirhe30xk4vg' start failed

Expected behavior

Che started and workspace created without issues

Runtime

• kubernetes (include output of kubectl version) Client Version: version.Info{Major:“1”, Minor:“12”, GitVersion:“v1.12.0”, GitCommit:“0ed33881dc4355495f623c6f22e7dd0b7632b7c0”, GitTreeState:“clean”, BuildDate:“2018-09-27T17:05:32Z”, GoVersion:“go1.10.4”, Compiler:“gc”, Platform:“linux/amd64”} Server Version: version.Info{Major:“1”, Minor:“13+”, GitVersion:“v1.13.10-r0-CCE2.0.26.B001”, GitCommit:“dd2361511301aa78e37385126753ad08eb0f616e”, GitTreeState:“clean”, BuildDate:“2019-09-09T19:19:38Z”, GoVersion:“go1.11.13”, Compiler:“gc”, Platform:“linux/amd64”}

Screenshots

Installation method

• chectl chectl server:start --platform=k8s --installer=helm --multiuser --domain=che-dev.esa-maap.org --k8spodwaittimeout=400000 --postgres-pvc-storage-class-name=“sata-iac-0b” --workspace-pvc-storage-class-name=“sata-iac-0b” --templates=/home/cloud/chectl-templates/DEV/templates

Environment

• Cloud Orange

Eclipse Che Logs

We are not able to find correct log

[sleshchenko]

I suppose it’s mainly caused by fact that all plugins we have are Che Theia compatible and they are not going to work with python.

[KOSSOKO]

@sleshchenko, please telll me if it if possible, to switch editor just by using UI provided for the same workspace ? I created on Redhat hosted eclipse che a workspace with the default stack python, then stop the workspace and switch Editor for Jupyter and even on Redhat, workspace not started.