fix: wait for volume to become available before cleanup deletion

[Autherain]

Fixes #123

Problem

When use_blockstorage_volume is enabled, the Cleanup() method in StepCreateVolume` attempts to delete the volume immediately after the has been created. However, Cinder may still be detaching the (status "uploading"), causing deletion to fail with a 400 error:

  Volume status must be available or error or error_restoring or or error_managing and must not be migrating, belong to a group, have snapshots, awaiting a transfer, be disassociated from snapshots after volume transfer.

This leaves orphaned volumes that require manual cleanup.

Fix

Added a call to the existing WaitForVolume() function before attempting deletion in Cleanup(). This polls until the volume reaches "available" or "error" status, which is consistent with how Run() already waits after volume creation.

If the wait fails, deletion is still attempted as a best-effort
fallback.

Testing

• go build ./... passes
• go test ./builder/openstack/... passes
• go fmt applied

[hashicorp-cla-app]

All committers have signed the CLA.

[Autherain]

Just a a friendly ping on this PR. It's been open since march 31 and has not been reviewed yet. It fixes a real issue we've been having around. Happy to make any changes needed !

[Autherain]

@tanmay-hc I forgot to ping you ;)

[tanmay-hc]

@anurag5sh Can you have a look?

[Copilot]

Pull request overview

This PR addresses orphaned Cinder volumes when use_blockstorage_volume is enabled by ensuring the volume is in a deletable state before Cleanup() attempts deletion, aligning cleanup behavior with the existing post-create wait in Run().

Changes:

• Add a pre-delete wait in StepCreateVolume.Cleanup() to wait for the volume to become available before attempting deletion.
• Improve cleanup logging to report wait failures while still attempting deletion as best-effort.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.