podvm: Cleanup leftover packer resources for AWS #592
Conversation
bpradipt
force-pushed
the
aws-packer-instances
branch
from
507babd to
55dc8c6
Compare
The cleanup script gets the default VPC for the region and then cleans up any left over packer instances. The packer created instances have `packer` as prefix Signed-off-by: Pradipta Banerjee <[email protected]>
bpradipt
force-pushed
the
aws-packer-instances
branch
from
55dc8c6 to
c772fea
Compare
|
/retest |
|
@bpradipt: all tests passed! Full PR test history. Your PR dashboard. 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-sigs/prow repository. I understand the commands that are listed here. |
There was a problem hiding this comment.
Functional-wise it works well. There is the question on how hard would be to add custom tags rather than rely on default VPC and packer* prefix, but we can probably get away with documenting this at this point. The bigger question is whether it's likely people will specify their own VPC when using this operator, in such case we should respect that in the cleanup.
There was a problem hiding this comment.
Functional-wise it works well. There is the question on how hard would be to add custom tags rather than rely on default VPC and packer* prefix, but we can probably get away with documenting this at this point. The bigger question is whether it's likely people will specify their own VPC when using this operator, in such case we should respect that in the cleanup.
There was a problem hiding this comment.
LGTM, provided we mention the default vpc packer* cleanup in docs. Tested with a few manually created vms without the operator and works well unless it hits some issues (like multiple VMs with the same SG, then it fails, which might be a good idea as our packer VMs should never do that)
|
Added release note text in the related bug - https://issues.redhat.com/browse/KATA-3573 |
- Description of the problem which is fixed/What is the use case
During pod VM image build on AWS, if the job get's terminated abruptly then there will be left over packer resources.
- What I did
Added a cleanup script to execute as part of preStop hook for the image creation job to delete any leftover packer instances.
- How to verify it
Create kataconfig and watch for image creation job to start. Wait till it reaches the packer VM creation step. Delete the image creation pod and watch it's log to see the cleanup happening.
- Description for the changelog
Cleanup left over AWS packer resources