✨ Add reference to HostUpdatePolicy in Servicing.

[rhjanders]

What this PR does / why we need it:

This PR enables BMO to run Ironic servicing operations (such as applying firmware settings changes - or in the future firmware updates to already provisioned nodes). Servicing is an opt-in feature and is controlled by creation of a HostUpdatePolicy for a node with attributes indicating the desire to make changes to firmware configuration onReboot.

This is a partial implementation of https://github.com/metal3-io/metal3-docs/blob/main/design/baremetal-operator/host-live-updates.md (please note only firmware settings changes are currently supported, firmware update support will be added next).

[metal3-io-bot]

Skipping CI for Draft Pull Request.
If you want CI signal for your change, please convert it to an actual PR.
You can still manually trigger a test run with /test all

[iurygregory]

LGTM, thanks for working on it @rhjanders

[metal3-io-bot]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by:
Once this PR has been reviewed and has the lgtm label, please assign zaneb for approval. For more information see the Kubernetes Code Review Process.

The full list of commands accepted by this bot can be found here.

Needs approval from an approver in each of these files:

• OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[iurygregory]

LGTM

[zaneb]

The actual thing we want to check here is whether we need to write the BMH. Writes occur on line 1406 and line 1420, but dirty could be true if hfsDirty is true even if nothing is actually updated.

[zaneb]

I think this is too early to clear an error. We should do it after the check for provResult.ErrorMessage != "" around line 1418.

It's also arguably too late for an attempt to set the status to Servicing in the non-error case, because we'll still only write it once servicing has already started.

[rhjanders]

I changed the order as per the first line of the comment. Unsure about the second part - will check in with Dmitry.

[zaneb]

Remember that writes to the k8s API only happen after we return from this function. So if you want to put it into OperationalStatusServicing before servicing starts then you'd need to do something like:

    if info.host.Status.OperationalStatus != metal3api.OperationalStatusServicing {
        info.host.Status.OperationalStatus = metal3api.OperationalStatusServicing
        return actionUpdate{}
    }

Just changing the order has no effect in that respect.

[dtantsur]

If we do it this way, we'll lose the error information, and it will never be passed to prov.Service.

[dtantsur]

I stand corrected. If we don't call clearErrorWithStatus, we'll have ErrorType still set. We just need to remember to unset it after the successful call.

Maybe we need to return clearErrorWithStatus to where it was before, but also set the status to servicing explicitly in the way that Zane suggested?

[dtantsur]

nit: you no longer need to store this because you no longer clear ErrorType before calling Service.

[dtantsur]

You probably need dirty = true here

[dtantsur]

/close

Superseded by #2041

[metal3-io-bot]

@dtantsur: Closed this PR.

In response to this:

/close

Superseded by #2041

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.