Elon Cheng

> > What is your version of Kubesphere? An adaptation for containerd is available in kubespherev3.2.Each change in ClusterConfiguration will cause the ks-Installer to be reinstalled, but the original configuration...

This looks like a problem with fluentbit, it might be better to change the version of fluentbit, `kubesphere/fluent-bit:v1.8.11`. Do we have to change the version of fluentbit? @benjaminhuo

This is the reason for helm. The reason is that helm unloads each cr in an unordered way, so it causes a certain CR to get stuck. `fluent-operator` does the...

Yes, it might be possible to add a preprocessing mechanism to determine the order of uninstallation of helm, which might be a new feature.

Hi @jayenzo . Can you please enable debug log level and share the log? https://github.com/fluent/fluent-operator/blob/master/apis/fluentbit/v1alpha2/clusterfluentbitconfig_types.go#L66

You can debug it like this. You go inside a pod and see if you can access the host. If the network is normal, you can change the fluentbit image...

> @wenchajun > > kubectl exec -i -t -n kubesphere-logging-system fluentd-forward-0 -c fluentd -- sh -c "clear; (bash || ash || sh) /usr/bin $ nc -zv elasticsearch-tool.internal 9200 elasticsearch-tool.internal (192.168.34.202:9200)...

This is automatically generated. We use KubeBuilder to develop fluentBit-Operator.For details, please refer to https://book-v1.book.kubebuilder.io/basics/status_subresource.html. CRD generation tool will use the +kubebuilder:subresource:status annotation to enable status subresource in the CRD...

It doesn't usually cause this problem, can you show me the logs?

@519859716 Currently, No liveness probe added to deployment's YAML Are you interested in collaborating on this?