2022踩坑记录原创
# cordns域名解析alpine镜像不兼容问题
20220622
alpine版本兼容性问题会导致backend pod里cordns解析失败 docker镜像是使用alpine作为底包,如果版本不匹配会导致域名解析问题 https://stackoverflow.com/questions/65181012/does-alpine-have-known-dns-issue-within-kubernetes https://kubernetes.io/zh-cn/docs/tasks/administer-cluster/dns-debugging-resolution/ https://github.com/kubernetes-sigs/kind/issues/2418
# kata-monitor宿主机服务对kata-deploy的影响
重启服务器会导致kata-deploy起不来
# containerd重启不影响容器进程;docker也可以
# 重启服务器containerd启动超时问题
https://github.com/containerd/containerd/issues/5597
# golang mkdir 777 权限问题
https://github.com/golang/go/issues/15210 https://zhuanlan.zhihu.com/p/33692995
# k8s nodename指定节点时资源不足导致failed pod不断创建,如cpu/mem/gpu资源不足
pod.spec.nodeName将Pod直接调度到指定的Node节点上,会【跳过Scheduler的调度策略】,该匹配规则是【强制】匹配。可以越过Taints污点进行调度。
Node didn't have enough resource: cpu, requested: 100000, used: 1440, capacity: 7800
Events:
Type Reason Age From Message
---- ------ ---- ---- -------
Warning OutOfcpu 4m19s kubelet, tztest Node didn't have enough resource: cpu, requested: 100000, used: 1440, capacity: 7800
[root@tztest kbuser]# kubectl get pod -n hff
NAME READY STATUS RESTARTS AGE
hff-sts-0 1/1 Running 0 53d
web-demo-799fccc7dd-295bk 0/1 OutOfcpu 0 7s
web-demo-799fccc7dd-7z6fm 0/1 OutOfcpu 0 11s
web-demo-799fccc7dd-n6jzz 0/1 OutOfcpu 0 5s
web-demo-799fccc7dd-ph4vt 0/1 OutOfcpu 0 18s
web-demo-799fccc7dd-rqhk5 0/1 OutOfcpu 0 3s
web-demo-799fccc7dd-t74r7 0/1 OutOfcpu 0 13s
web-demo-799fccc7dd-tqhv5 0/1 OutOfcpu 0 9s
web-demo-799fccc7dd-xcldr 0/1 OutOfcpu 0 15s
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
上次更新: 2022/10/09, 17:42:01