Some tests perform significantly worse on single node azure:

• [sig-auth] ServiceAccounts ServiceAccountIssuerDiscovery should support OIDC discovery of service account issuer [Conformance] [Suite:openshift/conformance/parallel/minimal] [Suite:k8s]
• [sig-instrumentation][Late] Alerts shouldn't report any unexpected alerts in firing or pending state [Suite:openshift/conformance/parallel]
• [sig-imageregistry][Feature:ImageExtract] Image extract should extract content from an image [Skipped:Disconnected] [Suite:openshift/conformance/parallel]
• [sig-api-machinery][Feature:ServerSideApply] Server-Side Apply should work for apps.openshift.io/v1, Resource=deploymentconfigs [Suite:openshift/conformance/parallel]
• [sig-auth][Feature:SecurityContextConstraints] TestPodUpdateSCCEnforcement [Suite:openshift/conformance/parallel] openshift-tests azure,amd64,ovn,upgrade,upgrade-micro,single-node
• [sig-devex][Feature:Templates] templateinstance security tests should pass security tests [Suite:openshift/conformance/parallel]

See Sippy

Looking at the kinds of job runs, there are big clusters of failed tests, with error messages related to capacity ("the server is currently unable to handle the request"). I believe that the azure single node host simply can't handle the tests, at least not with the parallelism we have set.

In addition, these azure single node minor upgrade jobs have never passed, as far back as we have data in sippy (may 2022). (4.8 -> 4.9 and every minor upgrade since)

This job needs attention. You'll see several cards broken out for some of the top failures, and some general problems. Several have some analysis.