"[k8s.io] Monitoring should verify monitoring pods and all cluster nodes are available on influxdb using heapster" failed on rktnetes

[yifan-gu]

/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/test/e2e/monitoring.go:43
Expected error:
    <*errors.errorString | 0xc82083bd90>: {
        s: "expected to find one replica for RC or deployment with label influxGrafana but got 0",
    }
    expected to find one replica for RC or deployment with label influxGrafana but got 0
not to have occurred

cc @kubernetes/sig-node @euank @tmrts

[dchen1107]

Do we know the root cause of this?

[euank]

On our CI environment we're running with MONITORING=standalone when it expects influxdb. I can't recall the issue with influxdb that resulted in this (at some point it didn't come up).

I've verified that, at least now, it appears to come up correctly. I'll update our CI environment and verify that test passes with no impact on other tests and, if so, we can close this.

[dchen1107]

#27092 was merged, close the issue. Thanks!