provider/openstack: Fixes Boot From Volume

[jtopjian]

This commit fixes the previously broken "boot from volume" feature. It also
adds an acceptance test to ensure the feature continues to work.

The "delete_on_termination" option was also added.

Partly Fixes #3206

[jtopjian]

The root cause of this was the following:

if blockDeviceRaw, ok := d.Get("block_device").(map[string]interface{}); ok && blockDeviceRaw != nil {

It should have been

if blockDeviceRaw, ok := d.Get("block_device").([]interface{}); ok && blockDeviceRaw != nil {

or something similar.

Because of this, the block of code never worked and so it won't exist in anyone's currently deployed state files. I took this opportunity to convert the block_device type to TypeSet and used the same implementation as the AWS root_block_device parameter.

[jtopjian]

Maybe I've just been looking at this for too long, but it's starting to feel awkward to me. I don't like how I've had to make volume_id optional in order to accurately report the root/boot volume.

As an alternative, I have made a new parameter called attached_volumes that is entirely computed. This way, the user specifies the boot volume in block_device, supplemental volumes in volume, and all will be reported in attached_volumes.

The downside is that anyone using the current volume parameter for Computed information will need to make changes.

This variation can be seem here: jtopjian@1555964

What's the best thing to do in this situation?

[jtopjian]

Ignore the above. After reviewing the AWS implementation, I think the method used in this PR is correct. Leaving the above comment in case anyone would want to discuss, though.

I'm adding a second commit to this PR that does some supplemental volume cleanup and adds some extra tests. While the commit isn't required to fix the issue at hand, it does make the implementation cleaner and tries to prevents regression.

[jtopjian]

Closing in favor of #3232

[ghost]

I'm going to lock this issue because it has been closed for 30 days ⏳. This helps our maintainers find and focus on the active issues.

If you have found a problem that seems similar to this, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further.