Closed xrmzju closed 4 years ago
@gaocegege
I think we should.
/cc @johnugeorge
@xrmzju Are you interested in it? I'd appreciate if you could fix it.
I think we should.
/cc @johnugeorge
@xrmzju Are you interested in it? I'd appreciate if you could fix it.
i'll take it. and by the way, should we replace the godep with go mod?
@xrmzju We have the plan but I did not have the bandwidth for it. If you have time, please replace it with go mod. Thanks.
@xrmzju We have the plan but I did not have the bandwidth for it. If you have time, please replace it with go mod. Thanks.
sure, i have already replace it in my own repo. i'll open the pr soon.
Sounds good to me. we can do go mod separately. we have to do it for other repos too.
Agree.
we can do go mod separately. we have to do it for other repos too.
i create a configmap to customize the worker init container(add cpu/mem limits)
and the pytorch controller failed to create worker pod with errors:
and i googled it and found this issue, so i tried replace the "gopkg.in/yaml.v2" with "github.com/kubernetes-sigs/yaml" and everything works fine. should we replace it in upstream