New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Automated cherry pick of #52131 #52575 upstream release 1.8 #54023
Automated cherry pick of #52131 #52575 upstream release 1.8 #54023
Conversation
Hi @divyenpatel. Thanks for your PR. I'm waiting for a kubernetes member to verify that this patch is reasonable to test. If it is, they should reply with I understand the commands that are listed here. 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. I understand the commands that are listed here. |
/ok-to-test |
/lgtm |
/approve |
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: divyenpatel, jsafrane, luomiao Associated issue: 52131 The full list of commands accepted by this bot can be found here.
Needs approval from an approver in each of these OWNERS Files:
You can indicate your approval by writing |
#52131 contains significant feature development. Cherrypicks for patch releases should be restricted to bug fixes and security fixes. See https://github.com/kubernetes/sig-release/blob/master/ephemera/README.md#patch-release-manager |
@jpbetz bulk polling of volumes adds significant performance improvement for vsphere cloud provider. Interfaces were added long back in release 1.6 - #41306 and development made in the PR - #52131 is just an implementation for bulk polling and internal vsphere cloud provider library changes to support this feature. We have a customer ask (vmware-archive#175) to get this feature in official kubernetes release 1.8 / 1.7. |
@divyenpatel Let's get a clean test run and discuss. We'll also need an appropriate release-note. |
/test pull-kubernetes-unit |
@jpbetz Without bulk polling of volumes implementation for vSphere cloud provider, there is a huge performance overhead on vCenter from vSphere cloud provider. In the existing implementation, vSphere cloud provider makes 1 call to VC per volume to check if disk is attached. If there are many volumes, it would make multiple VC calls to check if the disk is attached. By default VolumesAreAttached() is called for every 1 minute. This would very much create a huge overhead on vCenter just to verify if disks are attached. Also, it would delay other operations - create/delete/attach/detach as VC would queue them to service each request. So, having this PR which implements BulkVerifyVolumes() API makes only single call to VC irrespective of the number of the volumes attached to kubernetes nodes. So, for every minute there would only one VC call. As @divyenpatel suggested, customers ran into these issues and I think we should have this fix cherry-picked into 1.8/1.7 release branches. |
cc @wojtek-t since this is also being proposed as a 1.7 patch |
Since this is isolated to vSphere cloud provider code and there is precedence for allowing larger fixes/changes than normally allowed for patches to cloud provider specific code, I'm okay with merging this for 1.8.3. After 1.8.2 is release tomorrow, I'll add |
@jpbetz Can you add |
/retest Review the full test history for this PR. |
/test all [submit-queue is verifying that this PR is safe to merge] |
Automatic merge from submit-queue. |
Cherry pick of #52131, #52575 on release-1.8
#52131: Implement bulk polling of volumes for vSphere
#52575: Unable to detach the vSphere volume from Powered off node
@BaluDontu
Release note: