Add support for kube_scheduler SLI metrics #15731
Conversation
b5f71d1 to
f43d0be
Compare
Codecov Report
Flags with carried forward coverage won't be shown. Click here to find out more. |
| except Exception as e: | ||
| self.log.debug("Unable to collect query slis endpoint: %s", e) | ||
| return False | ||
| self._slis_available = r.status_code != 404 and r.status_code != 403 |
There was a problem hiding this comment.
If this function is supposed to be called, then I think we want to ignore a 404, but at the very least log an error for a 403, as that means that their agent or environment is not properly configured and they should be made aware of that
|
|
||
| def assert_metric(name, **kwargs): | ||
| # Wrapper to keep assertions < 120 chars | ||
| aggregator.assert_metric(NAMESPACE + name, **kwargs) |
There was a problem hiding this comment.
| aggregator.assert_metric(NAMESPACE + name, **kwargs) | |
| aggregator.assert_metric(f"{NAMESPACE}.{name}", **kwargs) |
There was a problem hiding this comment.
Since we run tests in Python 2 I couldn't use f-strings, but I used format instead
| assert_metric('.slis.kubernetes_healthcheck', value=1, tags=['name:ping', 'type:healthz']) | ||
| assert_metric( | ||
| '.slis.kubernetes_healthchecks_total', value=2450, tags=['name:ping', 'status:success', 'type:healthz'] | ||
| ) |
There was a problem hiding this comment.
| assert_metric('.slis.kubernetes_healthcheck', value=1, tags=['name:ping', 'type:healthz']) | |
| assert_metric( | |
| '.slis.kubernetes_healthchecks_total', value=2450, tags=['name:ping', 'status:success', 'type:healthz'] | |
| ) | |
| assert_metric('slis.kubernetes_healthcheck', value=1, tags=['name:ping', 'type:healthz']) | |
| assert_metric( | |
| 'slis.kubernetes_healthchecks_total', value=2450, tags=['name:ping', 'status:success', 'type:healthz'] | |
| ) |
| CHECK_NAME = 'kube_scheduler' | ||
| NAMESPACE = 'kube_scheduler' |
There was a problem hiding this comment.
Nit: personally I think you can just remove NAMESPACE and use CHECK_NAME.
|
|
||
| @pytest.fixture() | ||
| def mock_metrics(): | ||
| f_name = os.path.join(os.path.dirname(__file__), 'fixtures', 'metrics_slis_1.27.3.txt') |
There was a problem hiding this comment.
Instead of os.path.dirname(__file__), you can call get_here(). Example:
integrations-core/kong/tests/common.py
Line 11 in 3567074
|
The |
1 similar comment
|
The |
4b479a2 to
563d9ef
Compare
What does this PR do?
This PR introduces two new scheduler metrics:
kube_scheduler.slis.kubernetes_healthcheckandkube_scheduler.slis.kubernetes_healthchecks_total.Motivation
Kubernetes v1.26 exposed a new
/metrics/slisendpoint (reference here). This PR adds support for capturing the new metrics exposed for the scheduler:kubernetes_healthcheckandkubernetes_healthcheck.Additional Notes
Review checklist (to be filled by reviewers)
qa/skip-qalabel if the PR doesn't need to be tested during QA.