cosa aliyun-replicate is not idempotent

[aaradhak]

With the recent 4.14-9.2 build 414.92.202309142337-0, the cloud-replicate job seems to due to image duplication error where it mentions that the destination image exists.

[2023-09-18T13:58:11.963Z] Copying image to region "cn-huhehaote": copying image: SDK.ServerError
[2023-09-18T13:58:11.963Z] ErrorCode: InvalidImageName.Duplicated
[2023-09-18T13:58:11.963Z] Recommend: https://api.aliyun.com/troubleshoot?q=InvalidImageName.Duplicated&product=Ecs
[2023-09-18T13:58:11.963Z] RequestId: DE8E3152-55F2-3502-81DE-631FAFE6ACF9
[2023-09-18T13:58:11.963Z] Message: The destination image is exist.

Ref to the build job - https://jenkins-rhcos.apps.ocp-virt.prod.psi.redhat.com/job/cloud-replicate/35/

[jlebon]

This is at its root a cosa issue, but issues can't be transferred across orgs so we can just leave it here.

To fix this, we need to implement idempotency in the Aliyun image copy path just like we do in the AWS image copy path (where we know to check first if the image already exists and skip it if so).

[travier]

@aaradhak Can you refile this in https://github.com/coreos/coreos-assembler ? Thanks

[aaradhak]

sure, I have opened this in coreos-assembler - coreos/coreos-assembler#3634

can we close this issue here?

[travier]

Thanks, closing.
/close

[openshift-ci]

@travier: Closing this issue.

In response to this:

Thanks, closing.
/close

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.