Open rads18 opened 4 years ago
How did you increase the limits? Did you increase the system limits as well?
yes Openwhisk was configured to the new limits if thats what you mean?
hello team, Can someone provide me insight if there is something we am doing wrong or this falls in as a bug? If it is possible to increase the default system limits on action I'd appreciate if you could share a link on steps (assuming it is documented) so we can investigate what we did wrong.
thank you!
Since in our downstream, we are running OpenWhisk with higher memory and time limits, I am suspicious the way you configured new limits.
Could you share the details on how did you configure them?
Hi style
The team member that admins that environment is OOO at the moment if you can share which file do you expect these changes to be made?
(the content of the file i have shared in my initial post) thanks
Those configurations conform to this file: https://github.com/apache/openwhisk/blob/master/common/scala/src/main/resources/application.conf#L437
And I think you can configure them with these ansible variables. https://github.com/apache/openwhisk/blob/05ed4e18eb53f5630735ffbe28726dce5a3481b5/ansible/roles/controller/tasks/deploy.yml#L212 https://github.com/apache/openwhisk/blob/05ed4e18eb53f5630735ffbe28726dce5a3481b5/ansible/roles/controller/tasks/deploy.yml#L216
Thank you style for your inputs, i will be testing these with the admin in next day or so after they return. i found the ansible variable quite interesting , are these being set for a play called by OpenWhisk(OW) admin privileges ? Is there a way for me to pass when calling the action from ansible? or action parameters at whisk?
Note: I am not the admin for OW but have complete access in my namepsace on OW.
Hi Style,
I checked with my admin and they did update the controller file and also see the new values when they describe the pod on openwhisk-controller.
I have attached configuration file with new values. We basically took the max to 1MB.
Please let me know your thoughts thanks
( i dont see a way to attach so including it in here) The file name is openwhisk-controller_0.yaml
`apiVersion: v1 kind: Pod metadata: annotations: kubernetes.io/psp: aaaa-vault-psp creationTimestamp: "2020-09-03T20:06:54Z" generateName: openwhisk-controller- labels: app: openwhisk-openwhisk chart: openwhisk-0.1.6 controller-revision-hash: openwhisk-controller-6f6587bfb8 heritage: Tiller name: openwhisk-controller release: openwhisk statefulset.kubernetes.io/pod-name: openwhisk-controller-0 name: openwhisk-controller-0 namespace: openwhisk ownerReferences:
hostname | cut -d'-' -f3
env:Those configurations are system-wide configurations and users can change action metadata within those limits. And your k8s configurations look fine on the surface. Are you still unable to update your action limits accordingly?
Just to clarify my initial post again the action allows me to set the new limits but fails with "Internal Whisk error", in env A. However, we did a fresh setup in another environment(B) recently where the new limits were setup during install and those work fine.
This means in existing env A we are missing some configuration file which does not allow the setup to work. We can reinstall the existing environment with new higher limits but i dont believe it to be a good solution, specially to reinstall for anytime the environment variables change.
Environment details:
*Kubernetes, kubernetes 1.13
Docker version 18.09.7
Steps to reproduce the issue:
LT: Updated
max: "5m"
max: "15m"
LT
std: "1m" memory: min: "128m"
LT: Updated
max: "512m"
max: "1024m"
LT
std: "256m"
Provide the expected results and outputs:
Provide the actual results and outputs:
Additional information you deem important: