TOC

前言

开始部署

由于 Apache bookkeeper(本文后续简称为BK)需要使用元数据服务中心,当前支持 Zookeeper 和 Etcd,本文将会分别介绍使用 zookeeper+bookkeeper 部署和 etcd+bookkeeper 的部署,值得注意的是本文介绍的都是单机部署而不是集群.

会介绍三种部署方式:

1. 单独部署 bookkeeper 和 zookeeper
   
2. 单独部署 bookkeeper 和 etcd
   
3. 部署一个嵌入式用于开发的本地 bookkeeper

zookeeper+bookkeeper

下载需要的软件包

1. zookeeper/etcd
   
2. bookkeeper
   

下载命令可以用下面的两条命令分别下载以及解压 zookeeper 和 bookkeeper:

wget https://dlcdn.apache.org/zookeeper/zookeeper-3.7.0/apache-zookeeper-3.7.0-bin.tar.gz  
tar -xvf apache-zookeeper-3.7.0-bin.tar.gz

wget https://dlcdn.apache.org/bookkeeper/bookkeeper-4.14.0/bookkeeper-server-4.14.0-bin.tar.gz  
tar -xvf bookkeeper-server-4.14.0-bin.tar.gz

启动zookeeper

$ cd apache-zookeeper-3.7.0-bin
$ bin/zkServer.sh start conf/zoo_sample.cfg  
ZooKeeper JMX enabled by default
Using config: conf/zoo_sample.cfg
Starting zookeeper ... STARTED

logs目录下会生成一个zookeeper-{username}-server-{hostname}.out这样格式的日志.我这里的日志名是zookeeper-oem-server-lan.out

查看日志是否正常:

$ tail -22f logs/zookeeper-oem-server-lan.out  
...
2021-12-26 21:41:41,749 [myid:] - INFO  [ProcessThread(sid:0 cport:2181)::PrepRequestProcessor@137] - PrepRequestProcessor (sid:0) started, reconfigEnabled=false
2021-12-26 21:41:41,750 [myid:] - INFO  [main:RequestThrottler@75] - zookeeper.request_throttler.shutdownTimeout = 10000
2021-12-26 21:41:41,759 [myid:] - INFO  [main:ContainerManager@84] - Using checkIntervalMs=60000 maxPerMinute=10000 maxNeverUsedIntervalMs=0
2021-12-26 21:41:41,759 [myid:] - INFO  [main:ZKAuditProvider@42] - ZooKeeper audit is disabled.

日志一切正常,也可以看下端口号是否正常被进程占用了,如果确定 Zookkeeper 是启动正常的可以省略这部分:

$ lsof -i:2181
COMMAND   PID USER   FD   TYPE DEVICE SIZE/OFF NODE NAME
java    24342  oem   58u  IPv6 173956      0t0  TCP *:2181 (LISTEN)

端口查看也是正常的.部署 Zookeeper 的部分就完成了.

启动bookkeeper

执行bin/bookkeeper bookie 1来启动 Bookkeeper 发现报错了:

2021-12-26 22:01:54,802 - INFO  - [main:BookieNettyServer@424] - Shutting down BookieNettyServer
2021-12-26 22:01:54,822 - ERROR - [main:Main@228] - Failed to build bookie server
org.apache.bookkeeper.bookie.BookieException$MetadataStoreException: Failed to get cluster instance id
	at org.apache.bookkeeper.discover.ZKRegistrationManager.getClusterInstanceId(ZKRegistrationManager.java:428)

这是因为需要先初始化元数据,执行bin/bookkeeper shell metaformat:

22:02:39,898 INFO  Socket connection established, initiating session, client: /192.168.3.123:46676, server: node123/192.168.3.123:2281
22:02:40,037 INFO  Session establishment complete on server node123/192.168.3.123:2281, sessionid = 0x100216aac630001, negotiated timeout = 16000
22:02:40,043 INFO  ZooKeeper client is connected now.
22:02:40,761 INFO  Successfully formatted BookKeeper metadata
22:02:40,892 INFO  Session: 0x100216aac630001 closed
22:02:40,892 INFO  EventThread shut down for session: 0x100216aac630001

然后再次尝试启动 Bookkeeper 就可以启动成功了:

执行:bin/bookkeeper bookie 1

2021-12-26 22:04:01,472 - INFO  - [BookieJournal-3181:JournalChannel@157] - Opening journal /tmp/bk-txn-dev-bin/current/17df9d84aba.txn
2021-12-26 22:04:01,732 - INFO  - [BookKeeperClientScheduler-OrderedScheduler-0-0:NetworkTopologyImpl@426] - Adding a new node: /default-rack/192.168.3.123:3181
2021-12-26 22:04:02,014 - INFO  - [vert.x-eventloop-thread-0:VertxHttpServer$2@84] - Starting Vertx HTTP server on port 8081
2021-12-26 22:04:02,315 - INFO  - [main:VertxHttpServer@91] - Vertx Http server started successfully
2021-12-26 22:04:02,315 - INFO  - [main:ComponentStarter@86] - Started component bookie-server.

localbookie方式部署

这种方式主要是用于本地开发环境.

微信公众号

扫描下面的二维码关注我们的微信公众号,第一时间查看最新内容。同时也可以关注我的Github,看看我都在了解什么技术,在页面底部可以找到我的Github。

温馨提示

本文还在持续创作中,如果你对本文主题感兴趣可以加我微信好友进行催更,博客下方可以找到我的微信联系方式 :)

---