Bo Zhang

``` root@salt-master:~/dev-infrastructure/hetzner-k8s/kubeone# k get node NAME STATUS ROLES AGE VERSION hetzner-k8s-control-plane-1 Ready control-plane,master 43m v1.23.12 hetzner-k8s-pool1-5f676f7d8c-49s9b Ready 38m v1.23.12 hetzner-k8s-pool1-5f676f7d8c-58hhw Ready 38m v1.23.12 hetzner-k8s-pool1-5f676f7d8c-vt9dg Ready 38m v1.23.12 hetzner-k8s-pool1-5f676f7d8c-x8d7n Ready 8m28s...

I used `initial_machinedeployment_replicas=10` to start the cluster and only 4 worker nodes became available. After ~30min, some timeouts happened and I saw some instances destroyed and re-created, then 1 more...

The node seems to be working after restarting `kubelet` service... ``` root@hetzner-k8s-pool1-5f676f7d8c-pmd8r:~# service kubelet restart root@hetzner-k8s-pool1-5f676f7d8c-pmd8r:~# service kubelet status ● kubelet.service - kubelet: The Kubernetes Node Agent Loaded: loaded (/etc/systemd/system/kubelet.service;...

Further investigation shows that sometimes `service kubelet restart` is not enough, it also needs an extra `systemctl daemon-reload`. Anyway, I think this is a bug to resolve, because this will...

请点击一下`Java的方法和字段`，看看还有哪个任务没有完成

@bot-gradle test and merge

@bot-gradle test and merge

We ended up with a cron job removing the nodes that don't join for a long time: ``` * * * * * flock -n /tmp/fix-machine-if-not-join.lockfile -c "python3 /root/dev-infrastructure/hetzner-k8s/fix-machine-if-not-join.py key=/root/dev-infrastructure/hetzner-k8s/id_rsa...

@davidmichaelkarr This issue is probably due to the usage of ``` afterEvaluate { tagletPath ((configurations.taglet.files) as File[]) } ``` You shouldn't use `afterEvaluate`. When you apply `gradle-aggregate-javadocs-plugin`, it will register...

Thanks @kiview for reply, I was on vacation in the past days. Now I seem to be able to narrow down the problematic test to a single test case, and...