Skip to content

Conversation

@cgwalters
Copy link
Member

This occurs in the currently artificial scenario of the RHCOS
upgrade tests from previous 4.2 Anaconda-based builds. This wouldn't
happen in practice because growfs is already keyed off only executing
on the Ignition boot, and there's no way we'd rebase this code and
release updated 4.2 bootimages. But, we already have similar exclusions in
the mount code, so let's do the same for growfs.

This occurs in the currently artificial scenario of the RHCOS
upgrade tests from previous 4.2 Anaconda-based builds.  This wouldn't
happen in practice because growfs is already keyed off only executing
on the Ignition boot, and there's no way we'd rebase this code and
release updated 4.2 bootimages.  But, we already have similar exclusions in
the mount code, so let's do the same for growfs.
@dustymabe
Copy link
Member

can we add a comment that gives guidance on when this can be removed? Seems like something we'll only need for a short period of time.

@cgwalters
Copy link
Member Author

can we add a comment that gives guidance on when this can be removed? Seems like something we'll only need for a short period of time.

It's not any different from

But, we already have similar exclusions in the mount code, so let's do the same for growfs.

I don't think it's worth trying to remove these anytime soon.

Copy link
Member

@jlebon jlebon left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is fine, but it's essentially RHCOS-specific implementation details leaking into FCOS. I think we should work towards supporting RHCOS in the run-upgrade suite instead (see coreos/coreos-assembler#1158) and drop that artificial test.

@dustymabe
Copy link
Member

dustymabe commented Mar 18, 2020

Similar to the other mount rules, suppress invocation if we detect
we are running from a legacy setup created by Anaconda.

This is good info, but would be better if we say something like what was the last release we used Anaconda for. Then it's real easy to know when we no longer need it.

@jlebon jlebon merged commit acea416 into coreos:testing-devel Mar 18, 2020
@cgwalters
Copy link
Member Author

This is fine, but it's essentially RHCOS-specific implementation details leaking into FCOS.

Not quite right?

I think we should work towards supporting RHCOS in the run-upgrade suite instead (see coreos/coreos-assembler#1158) and drop that artificial test.

This I obviously agree with! But while this case was "synthetic", others won't be I think.

This is good info, but would be better if we say something like what was the last release we used Anaconda for. Then it's real easy to know when we no longer need it.

I think we need to support in-place updates from RHCOS 4.2 until such time as we have bootimage updates for OpenShift 4.

So...I'm not seeing us removing this anytime soon. It's not a huge tech debt at least.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants