Common use cases for this webhook involve using Kubernetes's
generateName API to randomise resource names (this is a good
idea in Tekton pipelines, for example, where there are uniqueness
constraints. That means that the webhook metrics here end up with
very high cardinality, which makes Prometheus fall over. Even
without generateName, it is possible to shoot oneself in the foot.
This commit just removes the resource_name label altogether.
* do not record for empty metric config
* Revert "do not record for empty metric config"
This reverts commit 539a5e4dbb.
* add a comment
* fix typo
* fix tests
* revert
* revert tests
* revert
* fix conflicts
* one more test file
With this change folks will need to call `webhook.RegisterMetrics()` to register the opencensus view with the metrics for the webhook's StatsReporter. This is needed to avoid having `sharedmain` crashloop the activator due to linking multiple on-`init()` views that register metrics named `request_latencies`.
In general, I believe that we should move away from registering these views via `init()` and more towards the broader K8s MetricsProvider pattern.
* Add metrics to webhook package
Add metricstest package for shared helper functions for testing metrics
* Address PR
* Cleanup
* Fix import paths to fix build issues
* Fix import package path for test file
* Remove unnecessary formatting from error message
* Remove helper function only used once
* Add metric name to all error messages, make checkRowTags testing helper function
* Add common histogram bucket generator function to metrics package
* Fix CheckStatsNotReported check
* Reset metrics before each test so the tests are idempotent
* Make CheckStatsNotReported conditional clearer
James Turley
Dave Protasowski
Yanwei Guo
Matt Moore
Annie Fu