chengjoey
chengjoey
/test pull-kubernetes-integration-go-compatibility
> I agree that it's worth revisiting the pod admission logic in kubelet, but the particular case of rejecting pods due to changed labels seems working as intended. I don't...
+1 looks good to have this feature, currently can stop parallel tasks manually, but it would be better to add a feature switch to stop them automatically. @tektoncd/core-maintainers what do...
The reasons for OOM are as follows: 1. The socket collector will have a for loop that continuously starts the goroutine `handleMessages` https://github.com/kubernetes/ingress-nginx/blob/f19e9265b0ca266c7f2bc5e4d2ac137479e8b842/internal/ingress/metric/collectors/socket.go#L455-L465 2. latencyMetric.Observe(stats.Latency) deprecated latencyMetric will start goroutine...
> While a deletion of the deprecated metric is a right way, `--exclude-socket-metrics=nginx_ingress_controller_ingress_upstream_latency_seconds` works as a temporary solution for the problem. Yes, it can be solved by excluding it, but...
> @chengjoey can you kindly help and post explicitly the link to the doc that announces the deprecation of what you said "`latencyMetric"` . https://github.com/kubernetes/ingress-nginx/blob/b933310da5914adbf3f243b5cf902c245d5a0cef/docs/user-guide/monitoring.md?plain=1#L389-L391 The user-guide clearly states that...
@PamelaMei-SAP @DongWei-4 @SuperQ Should we still fix the timeout issue and be able to delete suckMount after `unix.Statfs` succeed?
look issue #2903, this seems to be the expected behavior
> I'm not sure if this behavior is expected, because in my test case the result can vary from _10.2s_ to an unknown amount of time, which in both cases...