etcd
用于分布式系统的最重要数据的分布式的、可靠的键值存储。
提示：
- 在阅读本文之前，建议先阅读 raft 算法
1. 简介
1.1. etcd 是什么？
ectd 是强一致的、分布式的键值存储，它提供一种可靠的方式，来存储分布式系统或机器集群需要访问的数据。它在网络分区期间优雅地处理 Leader 选举，并且可以容忍机器故障，即使是在 Leader 节点中。
1.2. 特性
1.2.1. 接口简单
使用标准的 HTTP 工具，比如 curl，读写值。
1.2.2. 键值存储
与标准文件系统一样，在层次结构组织的目录中，存储数据。
1.2.3. 监听变化
监听特定键或目录的变化，对值的变化做出反应。
1.2.4. 可选的 SSL 客户端证书认证
1.2.5. 以每个实例每秒 1000 次写入为基准
1.2.6. 用于密钥过期的可选 TTL
1.2.7. 通过 Raft 协议正确分发
2. 安装
关于从预构建二进制文件（pre-built binaries）或源码安装 etcd 的细节，请参阅官方文档。接下来讲述如何安装预构建二进制文件。
2.1. 安装预构建二进制文件
安装 etcd 的最简单方式是从预构建二进制文件：
解压归档文件。这将生成包含二进制文件的目录
将可执行的二进制文件添加到 PATH 中。比如，将二进制文件重命名和/或移动到 PATH 的某个目录中（比如
/usr/local/bin），或将前一步创建的目录添加到 PATH 中在 Shell 中测试，
etcd在 PATH 中：$ etcd --versionetcd Version: 3.5.0...
3. 如何设置 Demo ectd 集群
官网教程：https://etcd.io/docs/v3.5/tutorials/how-to-setup-cluster/
在 etcd 中设置集群指南
在每个 etcd 节点上，指定集群成员：
CLUSTER_STATE=newNAME_1=machine-1NAME_2=machine-2NAME_3=machine-3HOST_1=192.168.56.121HOST_2=192.168.56.122HOST_3=192.168.56.123CLUSTER=${NAME_1}=http://${HOST_1}:2380,${NAME_2}=http://${HOST_2}:2380,${NAME_3}=http://${HOST_3}:2380在每台机器上运行：
# 在第一台机器上运行THIS_NAME=${NAME_1}THIS_IP=${HOST_1}etcd --data-dir=data.etcd --name ${THIS_NAME} \ --initial-advertise-peer-urls http://${THIS_IP}:2380 --listen-peer-urls http://${THIS_IP}:2380 \ --advertise-client-urls http://${THIS_IP}:2379 --listen-client-urls http://${THIS_IP}:2379 \ --initial-cluster ${CLUSTER} \ --initial-cluster-state ${CLUSTER_STATE} --initial-cluster-token ${TOKEN}
# 在第二台机器上运行THIS_NAME=${NAME_2}THIS_IP=${HOST_2}etcd --data-dir=data.etcd --name ${THIS_NAME} \ --initial-advertise-peer-urls http://${THIS_IP}:2380 --listen-peer-urls http://${THIS_IP}:2380 \ --advertise-client-urls http://${THIS_IP}:2379 --listen-client-urls http://${THIS_IP}:2379 \ --initial-cluster ${CLUSTER} \ --initial-cluster-state ${CLUSTER_STATE} --initial-cluster-token ${TOKEN}
# 在第三台机器上运行THIS_NAME=${NAME_3}THIS_IP=${HOST_3}etcd --data-dir=data.etcd --name ${THIS_NAME} \ --initial-advertise-peer-urls http://${THIS_IP}:2380 --listen-peer-urls http://${THIS_IP}:2380 \ --advertise-client-urls http://${THIS_IP}:2379 --listen-client-urls http://${THIS_IP}:2379 \ --initial-cluster ${CLUSTER} \ --initial-cluster-state ${CLUSTER_STATE} --initial-cluster-token ${TOKEN}现在 etcd 集群已经就绪，使用 etcdctl 连接 etcd：
x
export ETCDCTL_API=3HOST_1=192.168.56.121HOST_2=192.168.56.122HOST_3=192.168.56.123ENDPOINTS=$HOST_1:2379,$HOST_2:2379,$HOST_3:2379
etcdctl --endpoints=$ENDPOINTS member list4. 如何访问 etcd
etcd 集群访问指南
put 命令用于写：
etcdctl --endpoints=$ENDPOINTS put foo "Hello World!"get 用于从 etcd 读：
etcdctl --endpoints=$ENDPOINTS get fooetcdctl --endpoints=$ENDPOINTS --write-out="json" get foo5. 如何根据前缀获取键
根据前缀提取 etcd 键的指南
etcdctl --endpoints=$ENDPOINTS put web1 value1etcdctl --endpoints=$ENDPOINTS put web2 value2etcdctl --endpoints=$ENDPOINTS put web3 value3
etcdctl --endpoints=$ENDPOINTS get web --prefix6. 如何删除键
描述删除 etcd 键的方式
etcdctl --endpoints=$ENDPOINTS put key myvalueetcdctl --endpoints=$ENDPOINTS del key
etcdctl --endpoints=$ENDPOINTS put k1 value1etcdctl --endpoints=$ENDPOINTS put k2 value2etcdctl --endpoints=$ENDPOINTS del k --prefix7. 如何在事务中进行多个写操作
进行事务性写入的指南
txn 将多个请求包装进单个事务：
xxxxxxxxxxetcdctl --endpoints=$ENDPOINTS put user1 badetcdctl --endpoints=$ENDPOINTS txn --interactive
compares:value("user1") = "bad"
success requests (get, put, delete):del user1
failure requests (get, put, delete):put user1 good8. 如何监视键
监视 etcd 键的指南
watch 用于获取未来变更的通知：
etcdctl --endpoints=$ENDPOINTS watch stock1etcdctl --endpoints=$ENDPOINTS put stock1 1000
etcdctl --endpoints=$ENDPOINTS watch stock --prefixetcdctl --endpoints=$ENDPOINTS put stock1 10etcdctl --endpoints=$ENDPOINTS put stock2 209. 如何创建租约
在 etcd 中创建租约（lease）的指南
lease 用于带 TTL 的写：
xxxxxxxxxxetcdctl --endpoints=$ENDPOINTS lease grant 300# lease 2be7547fbc6a5afa granted with TTL(300s)
etcdctl --endpoints=$ENDPOINTS put sample value --lease=2be7547fbc6a5afaetcdctl --endpoints=$ENDPOINTS get sample
etcdctl --endpoints=$ENDPOINTS lease keep-alive 2be7547fbc6a5afaetcdctl --endpoints=$ENDPOINTS lease revoke 2be7547fbc6a5afa# or after 300 secondsetcdctl --endpoints=$ENDPOINTS get sample10. 如何创建锁
在 etcd 中创建分布式锁的指南
lock 用于分布式锁：
etcdctl --endpoints=$ENDPOINTS lock mutex1
# another client with the same name blocksetcdctl --endpoints=$ENDPOINTS lock mutex111. 如何在 etcd 集群中实施主节点选举
在 etcd 集群中，实施主节点选举的指南
elect 用于主节点选举：
xxxxxxxxxxetcdctl --endpoints=$ENDPOINTS elect one p1
# another client with the same name blocksetcdctl --endpoints=$ENDPOINTS elect one p212. 如何检查集群状态
检查 etcd 集群状态的指南
为每台机器指定初始集群配置：
xxxxxxxxxxetcdctl --write-out=table --endpoints=$ENDPOINTS endpoint status
+------------------+------------------+---------+---------+-----------+------------+-----------+------------+--------------------+--------+| ENDPOINT | ID | VERSION | DB SIZE | IS LEADER | IS LEARNER | RAFT TERM | RAFT INDEX | RAFT APPLIED INDEX | ERRORS |+------------------+------------------+---------+---------+-----------+------------+-----------+------------+--------------------+--------+| 10.240.0.17:2379 | 4917a7ab173fabe7 | 3.5.0 | 45 kB | true | false | 4 | 16726 | 16726 | || 10.240.0.18:2379 | 59796ba9cd1bcd72 | 3.5.0 | 45 kB | false | false | 4 | 16726 | 16726 | || 10.240.0.19:2379 | 94df724b66343e6c | 3.5.0 | 45 kB | false | false | 4 | 16726 | 16726 | |+------------------+------------------+---------+---------+-----------+------------+-----------+------------+--------------------+--------|etcdctl --endpoints=$ENDPOINTS endpoint health
10.240.0.17:2379 is healthy: successfully committed proposal: took = 3.345431ms10.240.0.19:2379 is healthy: successfully committed proposal: took = 3.767967ms10.240.0.18:2379 is healthy: successfully committed proposal: took = 4.025451ms13. 如何保存数据库
关于 etcd 数据库快照的指南
snapshot 用于保存 etcd 数据库的时间点（point-in-time）快照：
快照只能从一个 etcd 节点请求，因此 --endpoints 标志应该只包含一个端点。
ENDPOINTS=$HOST_1:2379etcdctl --endpoints=$ENDPOINTS snapshot save my.db
Snapshot saved at my.dbxxxxxxxxxxetcdctl --write-out=table --endpoints=$ENDPOINTS snapshot status my.db
+---------+----------+------------+------------+| HASH | REVISION | TOTAL KEYS | TOTAL SIZE |+---------+----------+------------+------------+| c55e8b8 | 9 | 13 | 25 kB |+---------+----------+------------+------------+14. 如何添加和移除成员
处理 etcd 集群成员的指南
memeber 用于添加、移除、更新成员：
在每台机器上执行：
x
# For each machineTOKEN=my-etcd-token-1CLUSTER_STATE=newNAME_1=etcd-node-1NAME_2=etcd-node-2NAME_3=etcd-node-3HOST_1=192.168.56.121HOST_2=192.168.56.122HOST_3=192.168.56.123CLUSTER=${NAME_1}=http://${HOST_1}:2380,${NAME_2}=http://${HOST_2}:2380,${NAME_3}=http://${HOST_3}:2380在 node 1 上执行：
x
THIS_NAME=${NAME_1}THIS_IP=${HOST_1}etcd --data-dir=data.etcd --name ${THIS_NAME} \ --initial-advertise-peer-urls http://${THIS_IP}:2380 \ --listen-peer-urls http://${THIS_IP}:2380 \ --advertise-client-urls http://${THIS_IP}:2379 \ --listen-client-urls http://${THIS_IP}:2379 \ --initial-cluster ${CLUSTER} \ --initial-cluster-state ${CLUSTER_STATE} \ --initial-cluster-token ${TOKEN}在 node 2 上执行：
xxxxxxxxxxTHIS_NAME=${NAME_2}THIS_IP=${HOST_2}etcd --data-dir=data.etcd --name ${THIS_NAME} \ --initial-advertise-peer-urls http://${THIS_IP}:2380 \ --listen-peer-urls http://${THIS_IP}:2380 \ --advertise-client-urls http://${THIS_IP}:2379 \ --listen-client-urls http://${THIS_IP}:2379 \ --initial-cluster ${CLUSTER} \ --initial-cluster-state ${CLUSTER_STATE} \ --initial-cluster-token ${TOKEN}在 node3 上执行：
x
THIS_NAME=${NAME_3}THIS_IP=${HOST_3}etcd --data-dir=data.etcd --name ${THIS_NAME} \ --initial-advertise-peer-urls http://${THIS_IP}:2380 \ --listen-peer-urls http://${THIS_IP}:2380 \ --advertise-client-urls http://${THIS_IP}:2379 \ --listen-client-urls http://${THIS_IP}:2379 \ --initial-cluster ${CLUSTER} \ --initial-cluster-state ${CLUSTER_STATE} \ --initial-cluster-token ${TOKEN}然后使用 member remove 和 member add 命令替换成员：
x
# 获取 member IDexport ETCDCTL_API=3HOST_1=192.168.56.121HOST_2=192.168.56.122HOST_3=192.168.56.123etcdctl --endpoints=${HOST_1}:2379,${HOST_2}:2379,${HOST_3}:2379 member list
# 移除成员MEMBER_ID=300774bba1f38564etcdctl --endpoints=${HOST_1}:2379,${HOST_2}:2379,${HOST_3}:2379 \ member remove ${MEMBER_ID}
# 添加新成员 node 4export ETCDCTL_API=3NAME_1=etcd-node-1NAME_2=etcd-node-2NAME_4=etcd-node-4HOST_1=192.168.56.121HOST_2=192.168.56.122HOST_4=192.168.56.124 # 新成员etcdctl --endpoints=${HOST_1}:2379,${HOST_2}:2379 \ member add ${NAME_4} \ --peer-urls=http://${HOST_4}:2380下面，使用 --initial-cluster-state existing 标记启动新成员：
xxxxxxxxxx# [警告]如果从相同的磁盘空间启动新成员，# 确保移除旧成员的数据目录## 使用 “existing” 标记重启
TOKEN=my-etcd-token-1CLUSTER_STATE=existingNAME_1=etcd-node-1NAME_2=etcd-node-2NAME_4=etcd-node-4HOST_1=192.168.56.121HOST_2=192.168.56.122HOST_4=192.168.56.124 # 新成员CLUSTER=${NAME_1}=http://${HOST_1}:2380,${NAME_2}=http://${HOST_2}:2380,${NAME_4}=http://${HOST_4}:2380
THIS_NAME=${NAME_4}THIS_IP=${HOST_4}etcd --data-dir=data.etcd --name ${THIS_NAME} \ --initial-advertise-peer-urls http://${THIS_IP}:2380 \ --listen-peer-urls http://${THIS_IP}:2380 \ --advertise-client-urls http://${THIS_IP}:2379 \ --listen-client-urls http://${THIS_IP}:2379 \ --initial-cluster ${CLUSTER} \ --initial-cluster-state ${CLUSTER_STATE} \ --initial-cluster-token ${TOKEN}