快速開始：
 關於kafka說明可以參考：    0) kafka叢集的安裝, 準備三臺伺服器
    server1:192.168.0.1     server2:192.168.0.2     server3:192.168.0.3  
   2) 解壓            tar -zxvf kafka_2.10-0.10.2.0.tgz            mv kafka_2.10-0.10.2.0   /usr/local/kafka           mkdir  /usr/local/kafka/logs    3) 配置        修改/usr/local/kafka/config/server.properties, 其中broker.id, log.dirs, zookeeper.connect 必須根據實際情況進行修改，其他項根據需要自行斟酌。    如下：    #========================================== broker.id=1       #port=9092   #預設     num.network.threads=3       num.io.threads=8       socket.send.buffer.bytes=1048576       socket.receive.buffer.bytes=1048576       socket.request.max.bytes=104857600       log.dir=/usr/local/kafka/logs       num.partitions=2     num.recovery.threads.per.data.dir=1     log.retention.hours=168        log.segment.bytes=536870912      log.retention.check.interval.ms=300000   zookeeper.connect=192.168.0.1:2181,192.168.0.2:2181,192.168.0.3:2181/kafka     zookeeper.connection.timeout.ms=6000       #==========================================
   注：在zookeeper 建立一個kafka結點，這樣好管理。 4) 啟動kafka   針對伺服器server2,server3可以將server1複製到相應的目錄:      scp -r /usr/local/kafka [email protected]:/usr/local/      scp -r /usr/local/kafka [email protected]

:/usr/local/
  cd /usr/local/kafka   修改三臺機子：config/server.properties     server1 :       broker.id=1   server2 :       broker.id=2   server3 :       broker.id=3   啟動     /usr/local/kafka/bin/kafka-server-start.sh  /usr/local/kafka/config/server.properties &  
  檢視      lsof -i:9092  5) 建立Topic(包含一個分割槽，三個副本)   /usr/local/kafka/bin/kafka-topics.sh --create --zookeeper  192.168.0.1:2181,192.168.0.2:2181,192.168.0.3:2181/kafka  --replication-factor 3 --partitions 1 --topic  mytopic
  #刪除   #/usr/local/kafka/bin/kafka-topics.sh --delete  --zookeeper 192.168.0.1:2181,192.168.0.2:2181,192.168.0.3:2181/kafka  --topic  mytopic
6) 檢視topic情況   /usr/local/kafka/bin/kafka-topics.sh --list --zookeeper 192.168.0.1:2181,192.168.0.2:2181,192.168.0.3:2181/kafka
7) 建立傳送者（發）    ./bin/kafka-console-producer.sh --broker-list  192.168.0.1:9092,192.168.0.2:9092,192.168.0.3:9092  --topic mytopic     my msg1     my msg2     ^C 8)  建立消費者（收）    /usr/local/kafka/bin/kafka-console-consumer.sh --zookeeper  192.168.0.1:2181,192.168.0.2:2181,192.168.0.3:2181/kafka  --topic  mytopic  --from-beginning     my msg1     my msg2     ^C 9)  檢視描述
   /usr/local/kafka/bin/kafka-topics.sh --describe --zookeeper 192.168.0.1:2181,192.168.0.2:2181,192.168.0.3:2181/kafka  --topic mytopic
10)  殺掉server1上的broker   kill `lsof -i:9092 | sed -n '2p' | awk '{print $2}'` 

  kill `ps -ef | grep kafka.Kafka | grep -v grep | awk '{print $2}'`

=============================================================

常用問題：

=============================================================

1、 啟動失敗 正則解釋出錯

[2017-04-29 19:25:54,810] FATAL  (kafka.Kafka$)
java.lang.VerifyError: Uninitialized object exists on backward branch 162
Exception Details:
  Location:
    scala/util/matching/Regex.unapplySeq(Lscala/util/matching/Regex$Match;)Lscala/Option; @216: goto
  Reason:
    Error exists in the bytecode
  Bytecode:
    0x0000000: 2bc6 000a 2bb6 00ef c700 07b2 0052 b02b
    0x0000010: b600 f2b6 00f3 2ab6 0054 4d59 c700 0b57
    0x0000020: 2cc6 000d a700 c92c b600 f799 00c2 bb00
    0x0000030: 6059 b200 65b2 006a 043e c700 0501 bf1d
    0x0000040: 2bb6 00f8 b600 74b6 0078 2bba 0100 0000
    0x0000050: b200 93b6 0097 3a06 3a05 59c7 0005 01bf
    0x0000060: 3a04 1906 b200 93b6 009b a600 7619 04b2
    0x0000070: 00a0 a600 09b2 00a0 a700 71bb 00a2 5919
    0x0000080: 04b6 00a8 3a0b 2b19 0bb8 00fc b200 a0b7
    0x0000090: 00ac 3a07 1907 3a08 1904 b600 afc0 00a4
    0x00000a0: 3a09 1909 b200 a0a5 0034 bb00 a259 1909
    0x00000b0: b600 a83a 0b2b 190b b800 fcb2 00a0 b700
    0x00000c0: ac3a 0a19 0819 0ab6 00b3 190a 3a08 1909
    0x00000d0: b600 afc0 00a4 3a09 a7ff ca19 07a7 000c
    0x00000e0: 1904 1905 1906 b800 b9b7 00bc b02a 2bb6
    0x00000f0: 00ef b601 02b0                         
  Stackmap Table:
    same_frame(@11)
    same_frame(@15)
    full_frame(@39,{Object[#2],Object[#34],Object[#86]},{Object[#86]})
    same_frame(@46)
    full_frame(@63,{Object[#2],Object[#34],Object[#86],Integer},{Uninitialized[#46],Uninitialized[#46],Object[#98]})
    full_frame(@96,{Object[#2],Object[#34],Object[#86],Integer,Top,Object[#206],Object[#208]},{Uninitialized[#46],Uninitialized[#46],Object[#164]})
    full_frame(@123,{Object[#2],Object[#34],Object[#86],Integer,Object[#164],Object[#206],Object[#208]},{Uninitialized[#46],Uninitialized[#46]})
    full_frame(@162,{Object[#2],Object[#34],Object[#86],Integer,Object[#164],Object[#206],Object[#208],Object[#162],Object[#162],Object[#164],Top,Object[#4]},{Uninitialized[#46],Uninitialized[#46]})
    full_frame(@219,{Object[#2],Object[#34],Object[#86],Integer,Object[#164],Object[#206],Object[#208],Object[#162],Object[#162],Object[#164],Top,Object[#4]},{Uninitialized[#46],Uninitialized[#46]})
    full_frame(@224,{Object[#2],Object[#34],Object[#86],Integer,Object[#164],Object[#206],Object[#208]},{Uninitialized[#46],Uninitialized[#46]})
    full_frame(@233,{Object[#2],Object[#34],Object[#86],Integer,Object[#164],Object[#206],Object[#208]},{Uninitialized[#46],Uninitialized[#46],Object[#4]})
    full_frame(@237,{Object[#2],Object[#34],Object[#86]},{})

	at scala.collection.immutable.StringLike.r(StringLike.scala:287)
	at scala.collection.immutable.StringLike.r$(StringLike.scala:287)
	at scala.collection.immutable.StringOps.r(StringOps.scala:29)
	at scala.collection.immutable.StringLike.r(StringLike.scala:276)
	at scala.collection.immutable.StringLike.r$(StringLike.scala:276)
	at scala.collection.immutable.StringOps.r(StringOps.scala:29)
	at kafka.cluster.EndPoint$.<init>(EndPoint.scala:29)
	at kafka.cluster.EndPoint$.<clinit>(EndPoint.scala)
	at kafka.server.Defaults$.<init>(KafkaConfig.scala:63)
	at kafka.server.Defaults$.<clinit>(KafkaConfig.scala)
	at kafka.server.KafkaConfig$.<init>(KafkaConfig.scala:616)
	at kafka.server.KafkaConfig$.<clinit>(KafkaConfig.scala)
	at kafka.server.KafkaServerStartable$.fromProps(KafkaServerStartable.scala:28)
	at kafka.Kafka$.main(Kafka.scala:58)
	at kafka.Kafka.main(Kafka.scala)

查了一天都無助了，最後查原始碼，感覺好是jdk的問題，出錯的jdk是：jdk-8u20-linux-x64.tar.gz ，換成 jdk-8u131-linux-x64.tar.gz 就能正常啟動了。

2、doesn’t match stored brokerId 0 in meta.properties

錯誤的原因是log.dirs目錄下的meta.properties中配置的broker.id和配置目錄下的server.properties中的broker.id不一致了，解決問題的方法是將兩者修改一致後再重啟。

3、WARN Error while fetching metadata with correlation id  錯誤處理

[2016-10-14 06:36:18,401] WARN Error while fetching metadata with correlation id 0 : {test333=LEADER_NOT_AVAILABLE} (org.apache.kafka.clients.NetworkClient)
[2016-10-14 06:36:19,543] WARN Error while fetching metadata with correlation id 1 : {test333=LEADER_NOT_AVAILABLE} (org.apache.kafka.clients.NetworkClient)
[2016-10-14 06:36:19,680] WARN Error while fetching metadata with correlation id 2 : {test333=LEADER_NOT_AVAILABLE} (org.apache.kafka.clients.NetworkClient)
[2016-10-14 06:36:19,908] WARN Error while fetching metadata with correlation id 3 : {test333=LEADER_NOT_AVAILABLE} (org.apache.kafka.clients.NetworkClient)
[2016-10-14 06:36:20,116] WARN Error while fetching metadata with correlation id 4 : {test333=LEADER_NOT_AVAILABLE} (org.apache.kafka.clients.NetworkClient)
[2016-10-14 06:36:20,334] WARN Error while fetching metadata with correlation id 5 : {test333=LEADER_NOT_AVAILABLE} (org.apache.kafka.clients.NetworkClient)
[2016-10-14 06:36:20,505] WARN Error while fetching metadata with correlation id 6 : {test333=LEADER_NOT_AVAILABLE} (org.apache.kafka.clients.NetworkClient)
[2016-10-14 06:36:20,757] WARN Error while fetching metadata with correlation id 7 : {test333=LEADER_NOT_AVAILABLE} (org.apache.kafka.clients.NetworkClient)

解決：config/server.properties kafka必須配置hostname

4、java.io.IOException: Connection to 127.0.0.1:9092 (id: 0 rack: null) failed

java.io.IOException: Connection to 127.0.0.1:9092 (id: 0 rack: null) failed
    at kafka.utils.NetworkClientBlockingOps$.awaitReady$1(NetworkClientBlockingOps.scala:84)
    at kafka.utils.NetworkClientBlockingOps$.blockingReady$extension(NetworkClientBlockingOps.scala:94)
    at kafka.server.ReplicaFetcherThread.sendRequest(ReplicaFetcherThread.scala:244)
    at kafka.server.ReplicaFetcherThread.fetch(ReplicaFetcherThread.scala:234)
    at kafka.server.ReplicaFetcherThread.fetch(ReplicaFetcherThread.scala:42)
    at kafka.server.AbstractFetcherThread.processFetchRequest(AbstractFetcherThread.scala:118)
    at kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:103)
    at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:63)

未有停機的情況現出這個錯誤，解決： rm -rf /usr/local/kafka/logs/*