Setting metadata.generateName in kubernetes jobs (batch/v1 Job) allows to re-run jobs after applying the resource again without the need to delete the previous job [source].
Currently trying to set this field on a DaskJob resource results in missing metadata.name in object {kubernetes.dask.org/v1 DaskJob}
Why?
Having the name be randomised via generateName on a DaskJob resource would allow users to rerun those jobs without the need to delete the completed jobs.
In my use case I would like to deploy DaskJobs using ArgoCD as the k8s controller, generating the name on the fly would allow to rerun the job whenever a sync is triggered.
What?
Setting
metadata.generateName
in kubernetes jobs (batch/v1 Job
) allows to re-run jobs after applying the resource again without the need to delete the previous job [source]. Currently trying to set this field on a DaskJob resource results inmissing metadata.name in object {kubernetes.dask.org/v1 DaskJob}
Why?
Having the name be randomised via
generateName
on a DaskJob resource would allow users to rerun those jobs without the need to delete the completed jobs. In my use case I would like to deploy DaskJobs using ArgoCD as the k8s controller, generating the name on the fly would allow to rerun the job whenever a sync is triggered.How?
Following the
ObjectMeta
spec:Imo I don't see a reason to implement this feature in the python library, but maybe someone will prove me wrong 😄
As a sidenote, I'd be happy to work on this if determined to be useful.