search

kubesphere / ks-installer

Install KubeSphere on existing Kubernetes cluster
https://kubesphere.io
Apache License 2.0
532 stars 746 forks source link

KubeSphere 2.0.2 云平台安装测试结果 #23

Open rayzhou2017 opened 5 years ago

rayzhou2017 commented 5 years ago

Three install options:

  1. online
  2. offline
  3. KubeSphere only, i.e., install KubeSphere on existing k8s cluster or k8s-based distribution.

All are 64 bit OSes (x86)

qingcloud huawei cloud aliyun tencent cloud aws azure Google Cloud
Ubuntu 16.04 online allinone √ √2 √ √4 √
Ubuntu 16.04 online 1master 2nodes √ √2 √ √4 √
Ubuntu 18.04 online allinone √ √2 √ √ √ √
Ubuntu 18.04 online 1master 2nodes √ √2 √ √ √
CentOS 7.6 online allinone √ √ √ 5 √ √1
CentOS 7.6 online 1master 2nodes √ √ √5 √ √1
RHEL 7.4 online allinone √ √3
RHEL 7.4 online 1master 2nodes √ √3
  1. Image from AWS Marketplace
  2. huawei cloud安装报错及解决方法：https://github.com/kubesphere/ks-installer/issues/19 (pip安装) https://github.com/kubesphere/ks-installer/issues/20 (coredns无法启动)
  3. image from AWS 社区AMI，online失败，需要用offline安装包且注释掉ebtables https://github.com/kubesphere/ks-installer/issues/23#issuecomment-523239730
  4. tencent cloud中ubuntu16.04镜像较旧(ubuntu16.04.1)，可能会遇到 illegal instruction (core dump)，规避方法可参考：https://github.com/kubesphere/ks-installer/issues/21
  5. aliyun中centos7安装时可能会遇到pip版本问题，解决方法可参考：https://github.com/kubesphere/ks-installer/issues/16
mayuhope commented 5 years ago

腾讯云测试结果： 1、Ubuntu 16.04.1安装在线版单节点和多节点：均失败，issue：https://github.com/kubesphere/ks-installer/issues/21

2、Ubuntu 18.04.1安装在线版单节点：成功 Ubuntu 18.04.1安装在线版多节点：成功

3、Centos7.5、Centos 7.6 安装在线版单节点：成功 Centos 7.6安装在线版多节点：成功

4 、Ubuntu18.04.1 、Ubuntu 16.04.1、Ubuntu 14.04.1（不支持离线部署）

5、缺少RHEL 7.4这个操作系统版本

Forest-L commented 5 years ago

阿里云： 1、ubuntu16.04.6 单节点在线：OK；

2、ubuntu16.04.6 单节点/多节点 离线：OK， 需要添加对应系统版本软件依赖包，issue：https://github.com/kubesphere/ks-installer/issues/22

3、centos7.5 单节点/多节点：OK， 系统pip版本问题，处理方法参考issue：https://github.com/kubesphere/ks-installer/issues/16

4、centos7.6 单节点离线：OK 系统pip版本问题，处理方法参考issue：https://github.com/kubesphere/ks-installer/issues/16 需要添加对应系统版本软件依赖包，issue：https://github.com/kubesphere/kubesphere/issues/630

5、 rhel 7.4 单节点在线：Failure

6、ubuntu18.04.1单节点在线：OK

7、centos7.6单节点在线：OK 系统pip版本问题，处理方法参考issue：https://github.com/kubesphere/ks-installer/issues/16

8、centos7.6多节点在线：OK 系统pip版本问题，处理方法参考issue：https://github.com/kubesphere/ks-installer/issues/16

9、ubuntu16.04.6多节点在线：OK

10、ubuntu18.04多节点在线：OK

rayzhou2017 commented 5 years ago

阿里云： centos7.5和centos7.6 单节点/多节点offline 替换了centos-7.5-amd64.iso和centos-7.6-amd64.iso 成功 centos7.5和centos7.6 online 成功

ubuntu16.04.6 单节点/多节点offline 增加ubuntu-16.04.6-server-amd64.iso 成功 Ubuntu16.04.6 单节点 online 成功

这个测试我都看不懂，郭峰那个就很容易看懂

rayzhou2017 commented 5 years ago

aws rhel 7.4 online alline 出错 Screen Shot 2019-08-20 at 1 58 56 PM

pixiake commented 5 years ago

rhel 7.4由于官方yum源中没有安装脚本中指定的相关软件，所以导致安装失败

  1. 可以更换或者添加centos7的yum源尝试解决
  2. 下个版本中会对该问题做相关处理
pixiake commented 5 years ago

aws rhel 7.4 offline alline 出错 图片

该问题为当前系统中没有预装ebtables，而安装包中使用的软件依赖包也缺少ebtables，导致安装失败

临时解决办法： vi ../k8s/roles/kubernetes/preinstall/vars/redhat.yml 注释掉ebtables

ubuntu的话修改ubuntu.yml

或者单独下载ebtables安装

Forest-L commented 5 years ago

vmware和各个平台机器对比： 1、在青云和阿里云centos7.6 离线install版本可以正常部署； 2、而同样的离线安装包在vmware机器上还报有依赖问题； vmware弄的机器，开发者很难有8核以上的机器。

hing007210 commented 5 years ago

TASK [ks-devops/ks-devops : OpenPitrix | Waiting for openpitrix-db] *** Tuesday 10 September 2019 11:29:55 +0800 (0:00:00.396) 0:04:18.141 * FAILED - RETRYING: OpenPitrix | Waiting for openpitrix-db (15 retries left). FAILED - RETRYING: OpenPitrix | Waiting for openpitrix-db (14 retries left). FAILED - RETRYING: OpenPitrix | Waiting for openpitrix-db (13 retries left). FAILED - RETRYING: OpenPitrix | Waiting for openpitrix-db (12 retries left). FAILED - RETRYING: OpenPitrix | Waiting for openpitrix-db (11 retries left). FAILED - RETRYING: OpenPitrix | Waiting for openpitrix-db (10 retries left). FAILED - RETRYING: OpenPitrix | Waiting for openpitrix-db (9 retries left). FAILED - RETRYING: OpenPitrix | Waiting for openpitrix-db (8 retries left). FAILED - RETRYING: OpenPitrix | Waiting for openpitrix-db (7 retries left). FAILED - RETRYING: OpenPitrix | Waiting for openpitrix-db (6 retries left). FAILED - RETRYING: OpenPitrix | Waiting for openpitrix-db (5 retries left). FAILED - RETRYING: OpenPitrix | Waiting for openpitrix-db (4 retries left). FAILED - RETRYING: OpenPitrix | Waiting for openpitrix-db (3 retries left). FAILED - RETRYING: OpenPitrix | Waiting for openpitrix-db (2 retries left). FAILED - RETRYING: OpenPitrix | Waiting for openpitrix-db (1 retries left). fatal: [master1]: FAILED! => {"attempts": 15, "changed": true, "cmd": "/usr/local/bin/kubectl -n openpitrix-system get pod | grep openpitrix-db-deployment | awk '{print $3}'", "delta": "0:00:00.113061", "end": "2019-09-10 11:32:28.753923", "rc": 0, "start": "2019-09-10 11:32:28.640862", "stderr": "", "stderr_lines": [], "stdout": "Pending", "stdout_lines": ["Pending"]}

hing007210 commented 5 years ago

在安装时报错。

Forest-L commented 5 years ago

@hing007210
1、麻烦额外建一个issue哦， 2、kubectl get pod -n openpitrix-system看下该namespace的pod是否正常启动， 3、然后kubectl describe pod -n openpitrix-system kubectl get pod -n openpitrix-system|grep openpitrix-db-deployment|awk '{print $1}' 查看pod的日志

Forest-L commented 5 years ago

经测试，通过以下方法可安装最新版本的pip（pip-19.2.3）可规避pip安装相关问题：

https://github.com/kubesphere/ks-installer/issues/16 https://github.com/kubesphere/ks-installer/issues/19 

curl https://bootstrap.pypa.io/get-pip.py -o get-pip.py
sudo python get-pip.py
yeyouqun commented 5 years ago

我在安装时，出现这个问题： Wednesday 18 September 2019 01:24:28 -0400 (0:00:00.639) 0:02:32.879 *** fatal: [ks-allinone]: FAILED! => {"changed": true, "cmd": "/usr/local/bin/helm upgrade --install ks-sonarqube /etc/kubesphere/sonarqube/sonarqube-0.13.5.tgz -f /etc/kubesphere/sonarqube/custom-values-sonarqube.yaml --namespace kubesphere-devops-system", "delta": "0:00:00.203654", "end": "2019-09-18 01:24:29.367462", "msg": "non-zero return code", "rc": 1, "start": "2019-09-18 01:24:29.163808", "stderr": "Error: could not find a ready tiller pod", "stderr_lines": ["Error: could not find a ready tiller pod"], "stdout": "", "stdout_lines": []}

PLAY RECAP ** ks-allinone : ok=172 changed=7 unreachable=0 failed=1

superfrankiefan commented 5 years ago

AWS UBUNTU 16.04.6 LTS

install mode: Kubesphere all in one

Error Description:

Stucked in the following task:

TASK [download : Check the local digest sha256 corresponds to the given image tag] ** Wednesday 09 October 2019 03:08:42 +0000 (0:00:00.057) 0:08:20.102 *****

TASK [download : container_download | Download containers if pull is required or told to always pull (delegate)] **** Wednesday 09 October 2019 03:08:42 +0000 (0:00:00.053) 0:08:20.156 *****

TASK [download : container_download | Download containers if pull is required or told to always pull (all nodes)] *** Wednesday 09 October 2019 03:08:42 +0000 (0:00:00.036) 0:08:20.192 * FAILED - RETRYING: container_download | Download containers if pull is required or told to always pull (all nodes) (4 retries left). changed: [ks-allinone]

Cloud someone give some suggetions?

Thanks & Regards, Frankie

sheeplinjie commented 5 years ago

部署了快好几个小时了，镜像还没拉下来呢，太慢了 就没有个线上存镜像的国内的仓库吗

inix commented 5 years ago

centos 7.5本地离线安装报错：

Thursday 24 October 2019  12:21:25 +0800 (0:00:00.370)       0:00:53.959 ******
fatal: [ks-allinone]: FAILED! => {"changed": true, "cmd": "/usr/local/bin/helm upgrade --install ks-sonarqube /etc/kubesphere/sonarqube/sonarqube-0.13.5.tgz -f /etc/kubesphere/sonarqube/custom-values-sonarqube.yaml --namespace kubesphere-devops-system", "delta": "0:00:02.181980", "end": "2019-10-24 12:21:28.168526", "msg": "non-zero return code", "rc": 1, "start": "2019-10-24 12:21:25.986546", "stderr": "Error: UPGRADE FAILED: Get https://10.233.0.1:443/api/v1/namespaces/kube-system/configmaps?labelSelector=NAME%!D(MISSING)ks-sonarqube%!C(MISSING)OWNER%!D(MISSING)TILLER%!C(MISSING)STATUS%!D(MISSING)DEPLOYED: dial tcp 10.233.0.1:443: connect: no route to host", "stderr_lines": ["Error: UPGRADE FAILED: Get https://10.233.0.1:443/api/v1/namespaces/kube-system/configmaps?labelSelector=NAME%!D(MISSING)ks-sonarqube%!C(MISSING)OWNER%!D(MISSING)TILLER%!C(MISSING)STATUS%!D(MISSING)DEPLOYED: dial tcp 10.233.0.1:443: connect: no route to host"], "stdout": "", "stdout_lines": []}

还有这个问题：

hursday 24 October 2019  14:23:33 +0800 (0:00:00.821)       0:11:49.833 ******
FAILED - RETRYING: ks-alerting | Waiting for alerting-db-init (30 retries left).
FAILED - RETRYING: ks-alerting | Waiting for alerting-db-init (29 retries left).
FAILED - RETRYING: ks-alerting | Waiting for alerting-db-init (28 retries left).
FAILED - RETRYING: ks-alerting | Waiting for alerting-db-init (27 retries left).
FAILED - RETRYING: ks-alerting | Waiting for alerting-db-init (26 retries left).
FAILED - RETRYING: ks-alerting | Waiting for alerting-db-init (25 retries left).
FAILED - RETRYING: ks-alerting | Waiting for alerting-db-init (24 retries left).
FAILED - RETRYING: ks-alerting | Waiting for alerting-db-init (23 retries left).
FAILED - RETRYING: ks-alerting | Waiting for alerting-db-init (22 retries left).
FAILED - RETRYING: ks-alerting | Waiting for alerting-db-init (21 retries left).
FAILED - RETRYING: ks-alerting | Waiting for alerting-db-init (20 retries left).
FAILED - RETRYING: ks-alerting | Waiting for alerting-db-init (19 retries left).
FAILED - RETRYING: ks-alerting | Waiting for alerting-db-init (18 retries left).
FAILED - RETRYING: ks-alerting | Waiting for alerting-db-init (17 retries left).
FAILED - RETRYING: ks-alerting | Waiting for alerting-db-init (16 retries left).
FAILED - RETRYING: ks-alerting | Waiting for alerting-db-init (15 retries left).
FAILED - RETRYING: ks-alerting | Waiting for alerting-db-init (14 retries left).
AkideLiu commented 5 years ago

vSphere 6.7u2平台 CentOS7.7： 1.all-in-one在线部署 ：成功 image 2.在 Kubernetes 在线部署 KubeSphere ：成功 kubernetes version：v1.13.5 image

3.在 Kubernetes 在线部署 KubeSphere ：失败> https://github.com/kubesphere/ks-installer/issues/334 kubernetes version：v1.15.3 基于terraform-vsphere-kubespray部署的kubernetes https://github.com/sguyennet/terraform-vsphere-kubespray

4.all-in-one离线部署 ：失败 因为系统是CentOS7.7，离线安装包里面只有CentOS 7.4 7.5 查看了kubespaere提供的镜像，不是官方标准镜像，所以没有继续测试。

xianyuwu commented 5 years ago

百度智能云 ubuntu18.04 在线安装成功。 image image 因为BCC配置比较低，所以有很多服务没启用。

focusonit commented 4 years ago

在金山云上安装2.02离线版，操作系统centos7.4，在etcd步骤出错：

TASK [etcd : Gen_certs | target ca-certificate store file] *** Tuesday 17 December 2019 08:50:10 +0800 (0:00:00.119) 0:01:08.026 ** ok: [node1] ok: [node2]

TASK [etcd : Gen_certs | add CA to trusted CA dir] *** Tuesday 17 December 2019 08:50:10 +0800 (0:00:00.104) 0:01:08.130 ** fatal: [node2]: FAILED! => {"changed": false, "msg": "Source /etc/ssl/etcd/ssl/ca.pem not found"} fatal: [node1]: FAILED! => {"changed": false, "msg": "Source /etc/ssl/etcd/ssl/ca.pem not found"}

NO MORE HOSTS LEFT ***

p-null commented 4 years ago

Hi, coming from the english documentation. Would it be better if we can change the issue title and content to be in English?