kubeasz/docs/setup/02-install_etcd.md

116 lines
3.7 KiB
Markdown
Raw Normal View History

2018-10-12 13:30:19 +08:00
## 02-安装etcd集群
2017-11-30 22:23:19 +08:00
2019-06-09 10:58:01 +08:00
kuberntes 集群使用 etcd 存储所有数据,是最重要的组件之一,注意 etcd集群需要奇数个节点(1,3,5...)本文档使用3个节点做集群。
2017-11-30 22:23:19 +08:00
请在另外窗口打开[roles/etcd/tasks/main.yml](../../roles/etcd/tasks/main.yml) 文件,对照看以下讲解内容。
2017-11-30 22:44:45 +08:00
2021-01-20 17:42:46 +08:00
### 创建etcd证书
2017-11-30 22:23:19 +08:00
2021-01-20 17:42:46 +08:00
注意证书是在部署节点创建好之后推送到目标etcd节点上去的以增加ca证书的安全性
2021-01-20 17:42:46 +08:00
创建ectd证书请求 [etcd-csr.json.j2](../../roles/etcd/templates/etcd-csr.json.j2)
2017-11-30 22:23:19 +08:00
``` bash
{
"CN": "etcd",
"hosts": [
2021-01-20 17:42:46 +08:00
{% for host in groups['etcd'] %}
"{{ host }}",
{% endfor %}
"127.0.0.1"
2017-11-30 22:23:19 +08:00
],
"key": {
"algo": "rsa",
"size": 2048
},
"names": [
{
"C": "CN",
"ST": "HangZhou",
"L": "XS",
"O": "k8s",
"OU": "System"
}
]
}
```
2021-01-20 17:42:46 +08:00
+ etcd使用对等证书hosts 字段必须指定授权使用该证书的 etcd 节点 IP这里枚举了所有ectd节点的地址
2017-11-30 22:23:19 +08:00
### 创建etcd 服务文件 [etcd.service.j2](../../roles/etcd/templates/etcd.service.j2)
2017-11-30 22:23:19 +08:00
``` bash
[Unit]
Description=Etcd Server
After=network.target
After=network-online.target
Wants=network-online.target
Documentation=https://github.com/coreos
[Service]
Type=notify
2022-06-28 19:59:50 +08:00
WorkingDirectory={{ ETCD_DATA_DIR }}
2017-11-30 22:23:19 +08:00
ExecStart={{ bin_dir }}/etcd \
2021-01-20 17:42:46 +08:00
--name=etcd-{{ inventory_hostname }} \
--cert-file={{ ca_dir }}/etcd.pem \
--key-file={{ ca_dir }}/etcd-key.pem \
--peer-cert-file={{ ca_dir }}/etcd.pem \
--peer-key-file={{ ca_dir }}/etcd-key.pem \
2017-11-30 22:23:19 +08:00
--trusted-ca-file={{ ca_dir }}/ca.pem \
--peer-trusted-ca-file={{ ca_dir }}/ca.pem \
--initial-advertise-peer-urls=https://{{ inventory_hostname }}:2380 \
--listen-peer-urls=https://{{ inventory_hostname }}:2380 \
--listen-client-urls=https://{{ inventory_hostname }}:2379,http://127.0.0.1:2379 \
--advertise-client-urls=https://{{ inventory_hostname }}:2379 \
2017-11-30 22:23:19 +08:00
--initial-cluster-token=etcd-cluster-0 \
--initial-cluster={{ ETCD_NODES }} \
2021-01-20 17:42:46 +08:00
--initial-cluster-state={{ CLUSTER_STATE }} \
2022-06-28 19:59:50 +08:00
--data-dir={{ ETCD_DATA_DIR }} \
--wal-dir={{ ETCD_WAL_DIR }} \
2021-01-20 17:42:46 +08:00
--snapshot-count=50000 \
--auto-compaction-retention=1 \
--auto-compaction-mode=periodic \
2022-06-28 19:59:50 +08:00
--max-request-bytes=10485760 \
2021-01-20 17:42:46 +08:00
--quota-backend-bytes=8589934592
Restart=always
RestartSec=15
2017-11-30 22:23:19 +08:00
LimitNOFILE=65536
2021-01-20 17:42:46 +08:00
OOMScoreAdjust=-999
2017-11-30 22:23:19 +08:00
[Install]
WantedBy=multi-user.target
```
2022-06-28 19:59:50 +08:00
2017-11-30 22:23:19 +08:00
+ 完整参数列表请使用 `etcd --help` 查询
2021-01-20 17:42:46 +08:00
+ 注意etcd 即需要服务器证书也需要客户端证书为方便使用一个peer 证书代替两个证书
+ `--initial-cluster-state` 值为 `new` 时,`--name` 的参数值必须位于 `--initial-cluster` 列表中
2022-06-28 19:59:50 +08:00
+ `--snapshot-count` `--auto-compaction-retention` 一些性能优化参数请查阅etcd项目文档
+ 设置`--data-dir` 和`--wal-dir` 使用不同磁盘目录可以避免磁盘io竞争提高性能具体请参考etcd项目文档
2017-11-30 22:23:19 +08:00
### 验证etcd集群状态
+ systemctl status etcd 查看服务状态
+ journalctl -u etcd 查看运行日志
2017-11-30 22:44:45 +08:00
+ 在任一 etcd 集群节点上执行如下命令
2017-11-30 22:23:19 +08:00
``` bash
# 根据hosts中配置设置shell变量 $NODE_IPS
export NODE_IPS="192.168.1.1 192.168.1.2 192.168.1.3"
for ip in ${NODE_IPS}; do
2018-05-06 09:08:09 +08:00
ETCDCTL_API=3 etcdctl \
2017-11-30 22:23:19 +08:00
--endpoints=https://${ip}:2379 \
--cacert=/etc/kubernetes/ssl/ca.pem \
--cert=/etc/etcd/ssl/etcd.pem \
--key=/etc/etcd/ssl/etcd-key.pem \
endpoint health; done
```
预期结果:
``` text
https://192.168.1.1:2379 is healthy: successfully committed proposal: took = 2.210885ms
https://192.168.1.2:2379 is healthy: successfully committed proposal: took = 2.784043ms
https://192.168.1.3:2379 is healthy: successfully committed proposal: took = 3.275709ms
```
三台 etcd 的输出均为 healthy 时表示集群服务正常。
2017-12-18 17:35:51 +08:00
2021-01-20 17:42:46 +08:00
[后一篇](03-container_runtime.md)