1、所需文件一览 Compose 文件内容如下：
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 name: zookeeper-cluster-demo networks: zookeeper-net: name: zookeeper-cluster-net driver: bridge services: zookeeper-1: image: ${image} restart: ${restart} hostname: ${host_name}-1 container_name: ${container_name}-1 networks: - zookeeper-net ports: - 127.0 .0 .1 :2181:2181 volumes: - ./zoo.cfg:/conf/zoo.cfg - ./${container_name}-1/data:/data - ./${container_name}-1/datalog:/datalog - ./${container_name}-1/logs:/logs environment: ZOO_MY_ID: 1 zookeeper-2: image: ${image} restart: ${restart} hostname: ${host_name}-2 container_name: ${container_name}-2 networks: - zookeeper-net ports: - 127.0 .0 .1 :2182:2181 volumes: - ./zoo.cfg:/conf/zoo.cfg - ./${container_name}-2/data:/data - ./${container_name}-2/datalog:/datalog - ./${container_name}-2/logs:/logs environment: ZOO_MY_ID: 2 zookeeper-3: image: ${image} restart: ${restart} hostname: ${host_name}-3 container_name: ${container_name}-3 networks: - zookeeper-net ports: - 127.0 .0 .1 :2183:2181 volumes: - ./zoo.cfg:/conf/zoo.cfg - ./${container_name}-3/data:/data - ./${container_name}-3/datalog:/datalog - ./${container_name}-3/logs:/logs environment: ZOO_MY_ID: 3 zookeeper-4: image: ${image} restart: ${restart} hostname: ${host_name}-4 container_name: ${container_name}-4 networks: - zookeeper-net ports: - 127.0 .0 .1 :2184:2181 volumes: - ./zoo.cfg:/conf/zoo.cfg - ./${container_name}-4/data:/data - ./${container_name}-4/datalog:/datalog - ./${container_name}-4/logs:/logs environment: ZOO_MY_ID: 4 zookeeper-5: image: ${image} restart: ${restart} hostname: ${host_name}-5 container_name: ${container_name}-5 networks: - zookeeper-net ports: - 127.0 .0 .1 :2185:2181 volumes: - ./zoo.cfg:/conf/zoo.cfg - ./${container_name}-5/data:/data - ./${container_name}-5/datalog:/datalog - ./${container_name}-5/logs:/logs environment: ZOO_MY_ID: 5
zoo.cfg 文件内容如下：
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 dataDir=/data dataLogDir=/datalog tickTime=2000 initLimit=10 syncLimit=5 autopurge.snapRetainCount=3 autopurge.purgeInterval=0 maxClientCnxns=60 standaloneEnabled=true admin.enableServer=true admin.serverPort=8080 server.1=zoo-1:2888:3888;2181 server.2=zoo-2:2888:3888;2181 server.3=zoo-3:2888:3888;2181 server.4=zoo-4:2888:3888;2181 server.5=zoo-5:2888:3888;2181
.env 文件内容如下：
1 2 3 4 image=zookeeper:3.8.4-jre-17 host_name=zoo restart=unless-stopped container_name=zookeeper
2、解析 Compose 文件 使用 docker compose config 命令来验证 Compose 文件是否正确：
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 $ docker compose config name: zookeeper-cluster-demo services: zookeeper-1: container_name: zookeeper-1 environment: ZOO_MY_ID: "1" hostname: zoo-1 image: zookeeper:3.8.4-jre-17 networks: zookeeper-net: null ports: - mode: ingress host_ip: 127.0.0.1 target: 2181 published: "2181" protocol: tcp restart: unless-stopped volumes: - type: bind source: E:\docker-volumes\zookeeper-cluster\zoo.cfg target: /conf/zoo.cfg bind: create_host_path: true - type: bind source: E:\docker-volumes\zookeeper-cluster\zookeeper-1\data target: /data bind: create_host_path: true - type: bind source: E:\docker-volumes\zookeeper-cluster\zookeeper-1\datalog target: /datalog bind: create_host_path: true - type: bind source: E:\docker-volumes\zookeeper-cluster\zookeeper-1\logs target: /logs bind: create_host_path: true # ...... networks: zookeeper-net: name: zookeeper-cluster-net driver: bridge
解析结果分析：
services.volumes 配置中显示类型为 bind，为绑定挂载，将容器中的 /data、/datalog、/zoo.cfg 文件映射到宿主机。
services.ports 配置指定映射容器的 2181 端口到宿主机的 2181-2185 端口，且指定 IP 为 127.0.0.1，如果不指定该 IP，除了宿主机，其他可以访问宿主机的机器也可以访问到该容器，这样是不安全的。另外，如果只是其他容器需要连接到该容器，而宿主机不需要访问容器，则可以不发布容器端口，而是指定 networks 属性，将容器连接到同一网络，从而使容器互相通信。例如，在另一个服务中使用 ZooKeeper Compose 中定义的网络。
1 2 3 4 5 6 7 8 9 10 11 12 13 networks: kafka-net: name: kafka-cluster-net driver: bridge zookeeper-net: name: zookeeper-cluster-net external: true services: broker-1: networks: - kafka-net - zookeeper-net
2.1 卷的多种选项 2.1.1 绑定挂载 1 2 3 4 5 6 7 services: zookeeper-1: volumes: - ./zoo.cfg:/conf/zoo.cfg - ./${container_name}-5/data:/data - ./${container_name}-5/datalog:/datalog - ./${container_name}-5/logs:/logs
2.1.2 匿名卷 1 2 3 4 services: zookeeper-1: volumes: - ./zoo.cfg:/conf/zoo.cfg
2.1.3 命名卷 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 volumes: volume-1-data: name: zookeeper-cluster-demo-volume-1-data volume-1-datalog: name: zookeeper-cluster-demo-volume-1-datalog volume-1-logs: name: zookeeper-cluster-demo-volume-1-logs services: zookeeper-1: volumes: - ./zoo.cfg:/conf/zoo.cfg - volume-1-data:/data - volume-1-datalog:/datalog - volume-1-logs:/logs
3、创建并启动容器 使用 docker compose up 命令来启动容器：
1 2 3 4 5 6 7 8 $ docker compose up -d [+] Running 6/6 ✔ Network zookeeper-cluster-net Created 0.0s ✔ Container zookeeper-cluster-demo-zookeeper-4-1 Started 0.0s ✔ Container zookeeper-cluster-demo-zookeeper-3-1 Started 0.0s ✔ Container zookeeper-cluster-demo-zookeeper-1-1 Started 0.0s ✔ Container zookeeper-cluster-demo-zookeeper-2-1 Started 0.0s ✔ Container zookeeper-cluster-demo-zookeeper-5-1 Started 0.1s
-d 选项为后台启动（分离模式启动）。
使用 docker ps 命令来查看刚才启动的容器：
1 2 3 4 5 6 7 $ docker ps -a CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES fb9a577c7b64 zookeeper:3.8.4-jre-17 "/docker-entrypoint.…" About a minute ago Up About a minute 2888/tcp, 3888/tcp, 8080/tcp, 0.0.0.0:2185->2181/tcp zookeeper-cluster-demo-zookeeper-5-1 49a5b27dd23b zookeeper:3.8.4-jre-17 "/docker-entrypoint.…" About a minute ago Up About a minute 2888/tcp, 3888/tcp, 8080/tcp, 0.0.0.0:2183->2181/tcp zookeeper-cluster-demo-zookeeper-3-1 208e8651a00b zookeeper:3.8.4-jre-17 "/docker-entrypoint.…" About a minute ago Up About a minute 2888/tcp, 3888/tcp, 0.0.0.0:2181->2181/tcp, 8080/tcp zookeeper-cluster-demo-zookeeper-1-1 781c40747ce9 zookeeper:3.8.4-jre-17 "/docker-entrypoint.…" About a minute ago Up About a minute 2888/tcp, 3888/tcp, 8080/tcp, 0.0.0.0:2182->2181/tcp zookeeper-cluster-demo-zookeeper-2-1 049260f67598 zookeeper:3.8.4-jre-17 "/docker-entrypoint.…" About a minute ago Up About a minute 2888/tcp, 3888/tcp, 8080/tcp, 0.0.0.0:2184->2181/tcp zookeeper-cluster-demo-zookeeper-4-1
4、查看 ZooKeeper 状态 4.1 ZooKeeper CLI 查看状态 1 2 3 4 5 6 7 8 9 10 11 $ docker exec -it fb9a577c7b64 zkServer.sh status ZooKeeper JMX enabled by default Using config: /conf/zoo.cfg Client port found: 2181. Client address: localhost. Client SSL: false. Mode: leader $ docker exec -it 49a5b27dd23b zkServer.sh status ZooKeeper JMX enabled by default Using config: /conf/zoo.cfg Client port found: 2181. Client address: localhost. Client SSL: false. Mode: follower
输出结果中的 Mode，展示了节点的模式，可以看到上面的 Leader 和 Follower 节点
使用 ZooKeeper CLI 连接 ZooKeeper 集群
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 $ docker exec -it fb9a577c7b64 zkServer.sh status $ zkCli.sh -server localhost 2181 Connecting to localhost:2181 2024-07-29 15:38:44,597 [myid:] - INFO [main:o.a.z.Environment@98] - Client environment:zookeeper.version=3.8.4-9316c2a7a97e1666d8f4593f34dd6fc36ecc436c, built on 2024-02-12 22:16 UTC 2024-07-29 15:38:44,597 [myid:] - INFO [main:o.a.z.Environment@98] - Client environment:host.name=zoo-5 2024-07-29 15:38:44,597 [myid:] - INFO [main:o.a.z.Environment@98] - Client environment:java.version=17.0.11 2024-07-29 15:38:44,597 [myid:] - INFO [main:o.a.z.Environment@98] - Client environment:java.vendor=Eclipse Adoptium 2024-07-29 15:38:44,598 [myid:] - INFO [main:o.a.z.Environment@98] - Client environment:java.home=/opt/java/openjdk 2024-07-29 15:38:44,598 [myid:] - INFO [main:o.a.z.Environment@98] - Client environment:java.class.path=/apache-zookeeper-3.8.4-bin/bin/../zookeeper-server/target/classes:/apache-zookeeper-3.8.4-bin/bin/../build/classes:/apache-zookeeper-3.8.4-bin/bin/../zookeeper-server/target/lib/*.jar:/apache-zookeeper-3.8.4-bin/bin/../build/lib/*.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/zookeeper-prometheus-metrics-3.8.4.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/zookeeper-jute-3.8.4.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/zookeeper-3.8.4.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/snappy-java-1.1.10.5.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/slf4j-api-1.7.30.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/simpleclient_servlet-0.9.0.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/simpleclient_hotspot-0.9.0.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/simpleclient_common-0.9.0.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/simpleclient-0.9.0.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/netty-transport-native-unix-common-4.1.105.Final.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/netty-transport-native-epoll-4.1.105.Final.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/netty-transport-classes-epoll-4.1.105.Final.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/netty-transport-4.1.105.Final.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/netty-resolver-4.1.105.Final.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/netty-handler-4.1.105.Final.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/netty-common-4.1.105.Final.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/netty-codec-4.1.105.Final.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/netty-buffer-4.1.105.Final.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/metrics-core-4.1.12.1.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/logback-core-1.2.13.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/logback-classic-1.2.13.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/jline-2.14.6.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/jetty-util-ajax-9.4.53.v20231009.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/jetty-util-9.4.53.v20231009.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/jetty-servlet-9.4.53.v20231009.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/jetty-server-9.4.53.v20231009.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/jetty-security-9.4.53.v20231009.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/jetty-io-9.4.53.v20231009.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/jetty-http-9.4.53.v20231009.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/javax.servlet-api-3.1.0.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/jackson-databind-2.15.2.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/jackson-core-2.15.2.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/jackson-annotations-2.15.2.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/commons-io-2.11.0.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/commons-cli-1.5.0.jar:/apache-zookeeper-3.8.4-bin/bin/../lib/audience-annotations-0.12.0.jar:/apache-zookeeper-3.8.4-bin/bin/../zookeeper-*.jar:/apache-zookeeper-3.8.4-bin/bin/../zookeeper-server/src/main/resources/lib/*.jar:/conf: 2024-07-29 15:38:44,598 [myid:] - INFO [main:o.a.z.Environment@98] - Client environment:java.library.path=/usr/java/packages/lib:/usr/lib64:/lib64:/lib:/usr/lib 2024-07-29 15:38:44,598 [myid:] - INFO [main:o.a.z.Environment@98] - Client environment:java.io.tmpdir=/tmp 2024-07-29 15:38:44,598 [myid:] - INFO [main:o.a.z.Environment@98] - Client environment:java.compiler=<NA> 2024-07-29 15:38:44,598 [myid:] - INFO [main:o.a.z.Environment@98] - Client environment:os.name=Linux 2024-07-29 15:38:44,598 [myid:] - INFO [main:o.a.z.Environment@98] - Client environment:os.arch=aarch64 2024-07-29 15:38:44,598 [myid:] - INFO [main:o.a.z.Environment@98] - Client environment:os.version=6.6.22-linuxkit 2024-07-29 15:38:44,598 [myid:] - INFO [main:o.a.z.Environment@98] - Client environment:user.name=root 2024-07-29 15:38:44,598 [myid:] - INFO [main:o.a.z.Environment@98] - Client environment:user.home=/root 2024-07-29 15:38:44,598 [myid:] - INFO [main:o.a.z.Environment@98] - Client environment:user.dir=/apache-zookeeper-3.8.4-bin 2024-07-29 15:38:44,598 [myid:] - INFO [main:o.a.z.Environment@98] - Client environment:os.memory.free=118MB 2024-07-29 15:38:44,598 [myid:] - INFO [main:o.a.z.Environment@98] - Client environment:os.memory.max=256MB 2024-07-29 15:38:44,598 [myid:] - INFO [main:o.a.z.Environment@98] - Client environment:os.memory.total=128MB 2024-07-29 15:38:44,599 [myid:] - INFO [main:o.a.z.ZooKeeper@637] - Initiating client connection, connectString=localhost:2181 sessionTimeout=30000 watcher=org.apache.zookeeper.ZooKeeperMain$MyWatcher@7113b13f 2024-07-29 15:38:44,600 [myid:] - INFO [main:o.a.z.c.X509Util@78] - Setting -D jdk.tls.rejectClientInitiatedRenegotiation=true to disable client-initiated TLS renegotiation 2024-07-29 15:38:44,602 [myid:] - INFO [main:o.a.z.ClientCnxnSocket@239] - jute.maxbuffer value is 1048575 Bytes 2024-07-29 15:38:44,606 [myid:] - INFO [main:o.a.z.ClientCnxn@1747] - zookeeper.request.timeout value is 0. feature enabled=false Welcome to ZooKeeper! 2024-07-29 15:38:44,607 [myid:localhost:2181] - INFO [main-SendThread(localhost:2181):o.a.z.ClientCnxn$SendThread@1177] - Opening socket connection to server localhost/[0:0:0:0:0:0:0:1]:2181. 2024-07-29 15:38:44,608 [myid:localhost:2181] - INFO [main-SendThread(localhost:2181):o.a.z.ClientCnxn$SendThread@1179] - SASL config status: Will not attempt to authenticate using SASL (unknown error) 2024-07-29 15:38:44,610 [myid:localhost:2181] - INFO [main-SendThread(localhost:2181):o.a.z.ClientCnxn$SendThread@1013] - Socket connection established, initiating session, client: /[0:0:0:0:0:0:0:1]:37150, server: localhost/[0:0:0:0:0:0:0:1]:2181 JLine support is enabled 2024-07-29 15:38:44,621 [myid:localhost:2181] - INFO [main-SendThread(localhost:2181):o.a.z.ClientCnxn$SendThread@1453] - Session establishment complete on server localhost/[0:0:0:0:0:0:0:1]:2181, session id = 0x5000010d7780000, negotiated timeout = 30000 WATCHER:: WatchedEvent state:SyncConnected type:None path:null
4.2 通过 AdminServer 查看节点状态 Zookeeper 通过 AdminServer 提供了一种基于 HTTP 的接口，可以使用 curl 命令来获取节点的状态信息。AdminServer 默认运行在端口 8080（可以根据配置进行更改）。你可以通过访问特定的 URL 来获取节点的各种信息，包括状态、配置信息等。
在 zoo.cfg 文件中有两个相关配置：
1 2 admin.enableServer=true admin.serverPort=8080
执行结果如下：
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 $ curl http://localhost:8080/commands/stats { "version" : "3.8.4-9316c2a7a97e1666d8f4593f34dd6fc36ecc436c, built on 2024-02-12 22:16 UTC", "read_only" : false, "server_stats" : { "packets_sent" : 33, "packets_received" : 33, "fsync_threshold_exceed_count" : 0, "client_response_stats" : { "last_buffer_size" : 16, "min_buffer_size" : 16, "max_buffer_size" : 16 }, "provider_null" : false, "uptime" : 2292719, "server_state" : "leader", "outstanding_requests" : 0, "min_latency" : 0, "avg_latency" : 1.44, "max_latency" : 7, "data_dir_size" : 718, "log_dir_size" : 67108880, "last_processed_zxid" : 4294967306, "num_alive_client_connections" : 0, "auth_failed_count" : 0, "non_mtlsremote_conn_count" : 0, "non_mtlslocal_conn_count" : 0 }, "client_response" : { "last_buffer_size" : 16, "min_buffer_size" : 16, "max_buffer_size" : 16 }, "proposal_stats" : { "last_buffer_size" : 48, "min_buffer_size" : 48, "max_buffer_size" : 48 }, "node_count" : 5, "connections" : [ ], "secure_connections" : [ ], "command" : "stats", "error" : null }
除了 /commands/stats 接口外，还有其他的接口例如：
http://localhost:8080/commands/monitor：获取监控信息http://localhost:8080/commands/config：获取配置信息http://localhost:8080/commands/envi：获取环境信息
5、环境变量 如果 Compose 文件中未提供 zoo.cfg 文件，则容器启动将使用 ZooKeeper 推荐的默认值。可以使用如下的环境变量进行覆盖：
ZOO_TICK_TIME
ZOO_INIT_LIMIT
ZOO_SYNC_LIMIT
ZOO_MAX_CLIENT_CNXNS
ZOO_STANDALONE_ENABLED
ZOO_ADMINSERVER_ENABLED
ZOO_AUTOPURGE_PURGEINTERVAL
ZOO_AUTOPURGE_SNAPRETAINCOUNT
ZOO_4LW_COMMANDS_WHITELIST
5.1 高级配置 并非所有的 ZooKeeper 都通过 Docker 环境变量进行公开。提供的环境变量仅涵盖最小配置和一些经常更改的选项。
可以使用该配置将额外的配置参数添加到 ZooKeeper 的配置文件中。
例如：
1 $ docker run --name some-zookeeper --restart always -e ZOO_CFG_EXTRA="metricsProvider.className=org.apache.zookeeper.metrics.prometheus.PrometheusMetricsProvider metricsProvider.httpPort=7070" zookeeper
5.1.2 JVMFLAGS 该配置用来设置 JVM 属性。
下面的配置选择使用 Netty 而不是 NIO 作为服务器通信框架：
1 $ docker run --name some-zookeeper --restart always -e JVMFLAGS="-Dzookeeper.serverCnxnFactory=org.apache.zookeeper.server.NettyServerCnxnFactory" zookeeper
下面配置则用来修改 JVM 的堆最大值：
1 $ docker run --name some-zookeeper --restart always -e JVMFLAGS="-Xmx1024m" zookeeper
5.2 复制模式 5.2.1 ZOO_MY_ID 该配置的值必须在集群中是唯一的，范围应该在1到255之间，不要求连续。
可以通过在数据目录下创建一个名为 myid 的文件，文件中只包含一服务器 ID，这个 ID 要与配置文件中的 ID 一致。
注意 ：如果挂载卷后容器的 /data 路径下存在 myid 文件，则此配置将无效。
5.2.2 ZOO_SERVERS 该变量允许您指定 Zookeeper 集群的机器列表。
每个条目应按以下格式指定：server.id=<address1>:<port1>:<port2>[:role];[<client port address>:]<client port>
条目之间用空格分隔。
注意 ：如果挂载卷后容器的 /conf 路径下已经存在 zoo.cfg 文件，则此配置将无效。
6、其他配置 6.1 日志 可以修改 /conf/logback.xml 文件来覆盖默认的日志记录配置，将自定义配置文件作为卷装载到容器中。
6.2 数据存储 /data 和 /datalog 挂载卷，分别用于保存 ZooKeeper 的内存数据快照和数据库事务日志。
注意放置事务日志的位置，专用的事务日志设备是保持良好性能的关键。将日志放在繁忙的设备上会对性能产生不利影响。
7、关于 ZooKeeper 群组 为了保证高可用性，ZooKeeper 以集群方式运行。由于使用了再均衡算法 ，建议一个 ZooKeeper 集群应该包含奇数个节点。
只有当群组中的大多数节点处于可用状态时，ZooKeeper 才能处理外部请求。也就是说，一个包含3个节点的群组，允许1个节点失效，而一个包含5个节点的群组允许2两个节点失效。
假设有一个包含 5 个节点的群组，如果要对群组做一些如更换节点的修改，那么就需要依次重启每个节点。如果你的群组无法容忍多个节点失效，那么在维护过程中就存在风险。
不过，也不建议一个群组包含超过 7 个节点，因为 ZooKeeper 使用了一致性协议，节点过多则会降低整个群组的性能。
此外，如果由于客户端连接太多，5 个或 7 个节点仍然无法支撑负载，则可以考虑增加额外的观察者节点来分摊只读流量。
相关链接 zookeeper - Official Image | Docker Hub
OB links #Docker #ZooKeeper #注册中心 #微服务
