fix: update metric when there are zero disruption candidates

[jmdeal]

Fixes #N/A

Description
Moves the eligible node metric update to the top-level disruption controller from the individual consolidation implementations. This both reduces code duplication and, more importantly, fixes a bug where the karpenter_disruption_eligible_nodes metric is not updated if the number of candidates is zero. This results in the last non-zero number of candidates being reported indefinitely.

How was this change tested?
Tested in a personal cluster via the Karpenter AWS provider

By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.

[coveralls]

Pull Request Test Coverage Report for Build 8980552943

Details

• 47 of 48 (97.92%) changed or added relevant lines in 5 files are covered.
• 9 unchanged lines in 2 files lost coverage.
• Overall coverage decreased (-0.06%) to 78.758%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
pkg/test/expectations/expectations.go	26	27	96.3%

Files with Coverage Reduction	New Missed Lines	%
pkg/controllers/disruption/drift.go	2	89.29%
pkg/controllers/provisioning/scheduling/preferences.go	7	86.67%

Totals
Change from base Build 8975908762:	-0.06%
Covered Lines:	8335
Relevant Lines:	10583

---

💛 - Coveralls

[njtran]

I understand making this its own separate test is nice to test individual things, but I'd rather not add more tests (when we're already having to increase the timeouts). Can you just add this eligible metric check into the existing disruption tests? If you can do this for each of the disruption tests, it'll make sure that the metric is working properly in all the different ways we're testing the codepaths too.

[jmdeal]

I had a pretty large set of updates for the disruption suite in general in my original consolidation race condition fix PR that I'm going to incorporate into a new PR. The main change was a rework of how we handle faking the clock which significantly sped up the test suite (~5x speed improvement IIRC). I think with that coming as well we can justify a standalone test, but I could also see incorporating this elsewhere.

[njtran]

Synced offline, opting for followups to solve this problem

[jonathan-innis]

Nice work tracking this down! This is a great simplifying change and some solid testing added!

[njtran]

/lgtm
/approve

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: jmdeal, njtran

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

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [njtran]

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