community icon indicating copy to clipboard operation
community copied to clipboard

request for mxjob status bug fix

Open xcloud-carl opened this issue 3 years ago • 1 comments

Hi. I want to request for mxjob status logic using training operator. When I create a training job with a simple command, two job's condition status are created at the same time. In fact, the job is completed, but it seems be Running when I check using 'kubectl get mxjob' My guess is because the Running status is aligned to the bottom. Could you fix this issue as soon as possible? Please reply.

API Version : kubeflow.org/v1 Kubeflow Version : 1.5

Status: Conditions: Last Transition Time: 2022-07-22T06:46:55Z Last Update Time: 2022-07-22T06:46:55Z Message: MXJob mxtrain1 is created. Reason: MXJobCreated Status: True Type: Created Last Transition Time: 2022-07-22T06:53:06Z Last Update Time: 2022-07-22T06:53:06Z Message: MXJob mxtrain1 is successfully completed. Reason: MXJobSucceeded Status: True Type: Succeeded Last Transition Time: 2022-07-22T06:53:06Z Last Update Time: 2022-07-22T06:53:06Z Message: MXJob mxtrain1 is running. Reason: MXJobRunning Status: True Type: Running

$ kubectl get mxjobs -n admin-0001 NAME STATE AGE mxtrain1 Running 18d

xcloud-carl avatar Aug 10 '22 00:08 xcloud-carl

Please create the PR in https://github.com/kubeflow/training-operator

johnugeorge avatar Aug 10 '22 13:08 johnugeorge