2.1安装失败，log如下，

runzexia · 2019年11月12日

https://github.com/helm/helm/issues/6894
是 helm v2.16.0 的问题。尝试安装一个其他版本的helm，再继续安装

koy1619 · 2019年11月12日

runzexia
只有 helm v2.16.0 有问题吗？

runzexia · 2019年11月12日

koy1619 https://github.com/helm/helm/issues/6894#issuecomment-551903514
参考issue 2.14.3/2.15.2 是没问题的

koy1619 · 2019年11月12日

runzexia
helm应该没问题了，这边还是过不了

TASK [common : Kubesphere | Creating manifests] ********************************
changed: [localhost] => (item={u'name': u'custom-values-minio', u'file': u'custom-values-minio.yaml'})

TASK [common : Kubesphere | Deploy minio] **************************************
fatal: [localhost]: FAILED! => {"changed": true, "cmd": "/usr/local/bin/helm upgrade --install ks-minio /etc/kubesphere/minio-ha -f /etc/kubesphere/custom-values-minio.yaml --set fullnameOverride=minio --namespace kubesphere-system --wait --timeout 1800\n", "delta": "0:00:00.213862", "end": "2019-11-12 07:51:56.429823", "msg": "non-zero return code", "rc": 1, "start": "2019-11-12 07:51:56.215961", "stderr": "Error: UPGRADE FAILED: \"ks-minio\" has no deployed releases", "stderr_lines": ["Error: UPGRADE FAILED: \"ks-minio\" has no deployed releases"], "stdout": "", "stdout_lines": []}
...ignoring

TASK [common : debug] **********************************************************
ok: [localhost] => {
    "msg": [
        "1. check the storage configuration and storage server", 
        "2. execute 'helm del --purge ks-minio'", 
        "3. Restart the installer pod in kubesphere-system namespace"
    ]
}

TASK [common : fail] ***********************************************************
fatal: [localhost]: FAILED! => {"changed": false, "msg": "It is suggested to refer to the above methods for troubleshooting problems ."}

PLAY RECAP *********************************************************************
localhost                  : ok=24   changed=18   unreachable=0    failed=1    skipped=66   rescued=0    ignored=1   


^C
[root@k8s-master deploy]# helm version
Client: &version.Version{SemVer:"v2.15.2", GitCommit:"8dce272473e5f2a7bf58ce79bb5c3691db54c96b", GitTreeState:"clean"}
Server: &version.Version{SemVer:"v2.15.2", GitCommit:"8dce272473e5f2a7bf58ce79bb5c3691db54c96b", GitTreeState:"clean"}
[root@k8s-master deploy]#

runzexia · 2019年11月12日

koy1619 按日志返回执行下卸载再继续安装

koy1619 · 2019年11月12日

runzexia

kubectl  delete -f .
sh ../scripts/kubesphere-delete.sh 
kubectl  apply -f .

还是上面那条报错！

runzexia · 2019年11月12日

koy1619 执行 ‘helm del –purge ks-minio’

koy1619 · 2019年11月12日

runzexia

[root@k8s-master deploy]# helm del --purge ks-minio 
release "ks-minio" deleted
[root@k8s-master deploy]#

重新卸载部署；这段报错过去了，谢谢大佬！

Hhetao · 2019年11月25日

koy1619 大佬，我的报错跟你一样，用kubectl get pod ks-installer-7987c659d6-vwl2t -n kubesphere-system -o yaml | kubectl replace –force -f - 这条命令重启installer后还是不行。
# helm version Client: &version.Version{SemVer:"v2.14.3", GitCommit:"0e7f3b6637f7af8fcfddb3d2941fcc7cbebb0085", GitTreeState:"clean"} Server: &version.Version{SemVer:"v2.14.3", GitCommit:"0e7f3b6637f7af8fcfddb3d2941fcc7cbebb0085", GitTreeState:"clean"}
系统版本centos 7.6，存储用的nfs

rysinal · 2019年12月4日

hetao runzexia 我的也是，centos 7.5 nfs存储，然后helm为2.14.3，安装2.1版本ks；试了两次都报错,卡的有点久了

helm del --purge ks-minio
kubectl get pod ks-installer-7987c659d6-sg8pv -n kubesphere-system -o yaml | kubectl replace --force -f -

TASK [common : Kubesphere | Deploy minio] **************************************

fatal: [localhost]: FAILED! => {"changed": true, "cmd": "/usr/local/bin/helm upgrade --install ks-minio /etc/kubesphere/minio-ha -f /etc/kubesphere/custom-values-minio.yaml --set fullnameOverride=minio --namespace kubesphere-system --wait --timeout 1800\n", "delta": "0:30:01.145146", "end": "2019-12-04 02:43:29.052773", "msg": "non-zero return code", "rc": 1, "start": "2019-12-04 02:13:27.907627", "stderr": "Error: release ks-minio failed: timed out waiting for the condition", "stderr_lines": ["Error: release ks-minio failed: timed out waiting for the condition"], "stdout": "Release \"ks-minio\" does not exist. Installing it now.", "stdout_lines": ["Release \"ks-minio\" doesnot exist. Installing it now."]}
...ignoring

TASK [common : debug] **********************************************************
ok: [localhost] => {
    "msg": [
        "1. check the storage configuration and storage server",
        "2. execute 'helm del --purge ks-minio'",
        "3. Restart the installer pod in kubesphere-system namespace"
    ]
}

TASK [common : fail] ***********************************************************
fatal: [localhost]: FAILED! => {"changed": false, "msg": "It is suggested to refer to the above methods for troubleshooting problems ."}

PLAY RECAP *********************************************************************
localhost                  : ok=24   changed=18   unreachable=0    failed=1    skipped=66   rescued=0    ignored=1

runzexia · 2019年12月4日

rysinal

首先删除 minio helm del --purge ks-minio
进入installer pod bash kubectl exec -n kubesphere-system ks-installer-xx -it sh
执行minio 安装命令 /usr/local/bin/helm upgrade --install ks-minio /etc/kubesphere/minio-ha -f /etc/kubesphere/custom-values-minio.yaml --set fullnameOverride=minio --namespace kubesphere-system --wait --timeout 1800
观察minio状态，排查为何出错

rysinal · 2019年12月4日

runzexia

[root@master ~]# helm del --purge ks-minio
release "ks-minio" deleted
[root@master ~]# kubectl get pods -n kubesphere-system
NAME                            READY   STATUS    RESTARTS   AGE
ks-installer-7987c659d6-rp9px   1/1     Running   0          40m
minio-make-bucket-job-rl9bw     1/1     Running   1          38m
openldap-0                      1/1     Running   0          14h
redis-5d4844b947-gzbsh          1/1     Running   0          14h
[root@master ~]# kubectl exec -n kubesphere-system ks-installer-7987c659d6-rp9px -it sh
/ # /usr/local/bin/helm upgrade --install ks-minio /etc/kubesphere/minio-ha -f /etc/kubesphere/custom-values-minio.yaml --set fullnameOverride=minio --namespace kubesphere
-system --wait --timeout 1800
Release "ks-minio" does not exist. Installing it now.
Error: jobs.batch "minio-make-bucket-job" already exists
/ #

rysinal · 2019年12月4日

rysinal

Events:
  Type     Reason       Age                From               Message
  ----     ------       ----               ----               -------
  Normal   Scheduled    50m                default-scheduler  Successfully assigned kubesphere-system/minio-make-bucket-job-rl9bw to node2
  Normal   Pulling      50m                kubelet, node2     Pulling image "minio/mc:RELEASE.2019-08-07T23-14-43Z"
  Normal   Pulled       49m                kubelet, node2     Successfully pulled image "minio/mc:RELEASE.2019-08-07T23-14-43Z"
  Normal   Created      11m (x3 over 49m)  kubelet, node2     Created container minio-mc
  Normal   Started      11m (x3 over 49m)  kubelet, node2     Started container minio-mc
  Normal   Pulled       11m (x2 over 30m)  kubelet, node2     Container image "minio/mc:RELEASE.2019-08-07T23-14-43Z" already present on machine
  Warning  FailedMount  1s (x12 over 11m)  kubelet, node2     MountVolume.SetUp failed for volume "minio-configuration" : [configmap "minio" not found, secret "minio" notfound]

rysinal · 2019年12月4日

看日志是存储问题，现在在每个节点安装了yum install -y nfs-utils
runzexia 大佬，现在报错为新的

TASK [common : Kubesphere | Deploy minio] **************************************
fatal: [localhost]: FAILED! => {"changed": true, "cmd": "/usr/local/bin/helm upgrade --install ks-minio /etc/kubesphere/minio-ha -f /etc/kubesphere/custom-values-minio.yaml --set fullnameOverride=minio --namespace kubesphere-system --wait --timeout 1800\n", "delta": "0:00:23.077751", "end": "2019-12-04 05:10:12.887874", "msg": "non-zero return code", "rc": 1, "start": "2019-12-04 05:09:49.810123", "stderr": "Error: jobs.batch \"minio-make-bucket-job\" already exists", "stderr_lines": ["Error: jobs.batch \"minio-make-bucket-job\" already exists"], "stdout": "Release \"ks-minio\" does not exist. Installing it now.", "stdout_lines": ["Release \"ks-minio\" does not exist. Installing it now."]}

删除了minio 的deveop和pod，包括job的pod，然后重新安装也是这样报错

kubectl delete deployment minio -n kubesphere-system
kubectl delete pod minio-8cd46c8d9-287gw -n kubesphere-system
kubectl delete pod minio-make-bucket-job-mvtzz -n kubesphere-system
helm del --purge ks-minio
kubectl get pod ks-installer-7987c659d6-ttprq -n kubesphere-system -o yaml | kubectl replace --force -f -

runzexia · 2019年12月4日

rysinal 把已经存在的job也删除一下

BBoBo-G · 2019年12月4日

rysinal 更改calio为flannel 一般都是因为底层网络不支持ipip协议影响k8s集群的dns解析所造成的

rysinal · 2019年12月4日

BoBo-G 大佬威武，已解决，是这个问题

Hhetao · 2019年12月11日

rysinal 查看一下异常的pod，describe或者看一下pod的日志，我失败的原因是写错了nfs的挂载目录

Raccoon-code · 2020年4月23日

rysinal 请问一下如何解决的，能知道一下吗

rysinal · 2020年4月23日

Raccoon-code 我的是更改calio为flannel

2.1安装失败，log如下，

runzexiaK零S

koy1619

runzexiaK零S

koy1619

runzexiaK零S

koy1619

runzexiaK零S

koy1619

HhetaoK零S

rysinalK零S

runzexiaK零S

rysinalK零S

rysinalK零S

rysinalK零S

runzexiaK零S

BBoBo-GK零S

rysinalK零S

HhetaoK零S

Raccoon-codeK零S

rysinalK零S