Kafka管理工具介绍

Kafka内部提供了许多管理脚本，这些脚本都放在$KAFKA_HOME/bin目录下，而这些类的实现都是放在源码的kafka/core/src/main/scala/kafka/tools/路径下。

Consumer Offset Checker

　　Consumer Offset Checker主要是运行kafka.tools.ConsumerOffsetChecker类，对应的脚本是kafka-consumer-offset-checker.sh，会显示出Consumer的Group、Topic、分区ID、分区对应已经消费的Offset、logSize大小，Lag以及Owner等信息。

如果运行kafka-consumer-offset-checker.sh脚本的时候什么信息都不输入，那么会显示以下信息：

[iteblog@www.iteblog.com /]$ bin/kafka-consumer-offset-checker.sh

Check the offset of your consumers.

Option                                  Description

------                                  -----------

--broker-info                           Print broker info

--group                                 Consumer group.

--help                                  Print this message.

--retry.backoff.ms <Integer>            Retry back-off to use for failed

                                          offset queries. (default: 3000)

--socket.timeout.ms <Integer>           Socket timeout to use when querying

                                          for offsets. (default: 6000)

--topic                                 Comma-separated list of consumer

                                          topics (all topics if absent).

--zookeeper                             ZooKeeper connect string. (default:

                                          localhost:2181)

我们根据提示，输入的命令如下：

[iteblog@www.iteblog.com /]$ bin/kafka-consumer-offset-checker.sh --zookeeper www.iteblog.com:2181 --topic test --group spark --broker-info

Group           Topic      Pid Offset          logSize         Lag             Owner

spark    test       0   34666914        34674392        7478            none

spark    test       1   34670481        34678029        7548            none

spark    test       2   34670547        34678002        7455            none

spark    test       3   34664512        34671961        7449            none

spark    test       4   34680143        34687562        7419            none

spark    test       5   34672309        34679823        7514            none

spark    test       6   34674660        34682220        7560            none

BROKER INFO

2 -> www.iteblog.com:9092

5 -> www.iteblog.com:9093

4 -> www.iteblog.com:9094

7 -> www.iteblog.com:9095

1 -> www.iteblog.com:9096

3 -> www.iteblog.com:9097

6 -> www.iteblog.com:9098

Dump Log Segment

　　有时候我们需要验证日志索引是否正确，或者仅仅想从log文件中直接打印消息，我们可以使用kafka.tools.DumpLogSegments类来实现，先来看看它需要的参数：

[iteblog@www.iteblog.com /]$ bin/kafka-run-class.sh kafka.tools.DumpLogSegments

Parse a log file and dump its contents to the console, useful for debugging a seemingly corrupt log segment.

Option                                  Description

------                                  -----------

--deep-iteration                        if set, uses deep instead of shallow

                                          iteration

--files <file1, file2, ...>             REQUIRED: The comma separated list of

                                          data and index log files to be dumped

--key-decoder-class                     if set, used to deserialize the keys.

                                          This class should implement kafka.

                                          serializer.Decoder trait. Custom jar

                                          should be available in kafka/libs

                                          directory. (default: kafka.

                                          serializer.StringDecoder)

--max-message-size <Integer: size>      Size of largest message. (default:

                                          5242880)

--print-data-log                        if set, printing the messages content

                                          when dumping data logs

--value-decoder-class                   if set, used to deserialize the

                                          messages. This class should

                                          implement kafka.serializer.Decoder

                                          trait. Custom jar should be

                                          available in kafka/libs directory.

                                          (default: kafka.serializer.

                                          StringDecoder)

--verify-index-only                     if set, just verify the index log

                                          without printing its content

　　很明显，我们在使用kafka.tools.DumpLogSegments的时候必须输入--files，这个参数指的就是Kafka中Topic分区所在的绝对路径。分区所在的目录由config/server.properties文件中log.dirs参数决定。比如我们想看/home/q/kafka/kafka_2.10-0.8.2.1/data/test-4/00000000000034245135.log日志文件的相关情况可以使用下面的命令：

[iteblog@www.iteblog.com /]$ bin/kafka-run-class.sh kafka.tools.DumpLogSegments --files /iteblog/data/test-4/00000000000034245135.log

Dumping /home/q/kafka/kafka_2.10-0.8.2.1/data/test-4/00000000000034245135.log

Starting offset: 34245135

offset: 34245135 position: 0 isvalid: true payloadsize: 4213 magic: 0 compresscodec: NoCompressionCodec crc: 865449274 keysize: 4213

offset: 34245136 position: 8452 isvalid: true payloadsize: 4657 magic: 0 compresscodec: NoCompressionCodec crc: 4123037760 keysize: 4657

offset: 34245137 position: 17792 isvalid: true payloadsize: 3921 magic: 0 compresscodec: NoCompressionCodec crc: 541297511 keysize: 3921

offset: 34245138 position: 25660 isvalid: true payloadsize: 2290 magic: 0 compresscodec: NoCompressionCodec crc: 1346104996 keysize: 2290

offset: 34245139 position: 30266 isvalid: true payloadsize: 2284 magic: 0 compresscodec: NoCompressionCodec crc: 1930558677 keysize: 2284

offset: 34245140 position: 34860 isvalid: true payloadsize: 268 magic: 0 compresscodec: NoCompressionCodec crc: 57847488 keysize: 268

offset: 34245141 position: 35422 isvalid: true payloadsize: 263 magic: 0 compresscodec: NoCompressionCodec crc: 2964399224 keysize: 263

offset: 34245142 position: 35974 isvalid: true payloadsize: 1875 magic: 0 compresscodec: NoCompressionCodec crc: 647039113 keysize: 1875

offset: 34245143 position: 39750 isvalid: true payloadsize: 648 magic: 0 compresscodec: NoCompressionCodec crc: 865445580 keysize: 648

offset: 34245144 position: 41072 isvalid: true payloadsize: 556 magic: 0 compresscodec: NoCompressionCodec crc: 1174686061 keysize: 556

offset: 34245145 position: 42210 isvalid: true payloadsize: 4211 magic: 0 compresscodec: NoCompressionCodec crc: 3691302513 keysize: 4211

offset: 34245146 position: 50658 isvalid: true payloadsize: 2299 magic: 0 compresscodec: NoCompressionCodec crc: 2367114411 keysize: 2299

offset: 34245147 position: 55282 isvalid: true payloadsize: 642 magic: 0 compresscodec: NoCompressionCodec crc: 4122061921 keysize: 642

offset: 34245148 position: 56592 isvalid: true payloadsize: 4211 magic: 0 compresscodec: NoCompressionCodec crc: 3257991653 keysize: 4211

offset: 34245149 position: 65040 isvalid: true payloadsize: 2278 magic: 0 compresscodec: NoCompressionCodec crc: 2103489307 keysize: 2278

offset: 34245150 position: 69622 isvalid: true payloadsize: 269 magic: 0 compresscodec: NoCompressionCodec crc: 792857391 keysize: 269

offset: 34245151 position: 70186 isvalid: true payloadsize: 640 magic: 0 compresscodec: NoCompressionCodec crc: 791599616 keysize: 640

可以看出，这个命令将Kafka中Message中Header的相关信息和偏移量都显示出来了，但是没有看到日志的内容，我们可以通过--print-data-log来设置。如果需要查看多个日志文件，可以以逗号分割。

导出Zookeeper中Group相关的偏移量

　　有时候我们需要导出某个Consumer group各个分区的偏移量，我们可以通过使用Kafka的kafka.tools.ExportZkOffsets类来满足。来看看这个类需要的参数：

[iteblog@www.iteblog.com /]$ bin/kafka-run-class.sh kafka.tools.ExportZkOffsets

Export consumer offsets to an output file.

Option                                  Description

------                                  -----------

--group                                 Consumer group.

--help                                  Print this message.

--output-file                           Output file

--zkconnect                             ZooKeeper connect string. (default:

                                          localhost:2181)

我们需要输入Consumer group，Zookeeper的地址以及保存文件路径：

[iteblog@www.iteblog.com /]$ bin/kafka-run-class.sh kafka.tools.ExportZkOffsets --group spark --zkconnect www.iteblog.com:2181 --output-file ~/offset

[iteblog@www.iteblog.com /]$ vim ~/offset

/consumers/spark/offsets/test/3:34846274

/consumers/spark/offsets/test/2:34852378

/consumers/spark/offsets/test/1:34852360

/consumers/spark/offsets/test/0:34848170

/consumers/spark/offsets/test/6:34857010

/consumers/spark/offsets/test/5:34854268

/consumers/spark/offsets/test/4:34861572

注意，--output-file参数必须在指定，否则会出错。

通过JMX获取metrics信息

　　我们可以通过kafka.tools.JmxTool类打印出Kafka相关的metrics信息。

[iteblog@www.iteblog.com /]$ bin/kafka-run-class.sh kafka.tools.JmxTool

Dump JMX values to standard output.

Option                                  Description

------                                  -----------

--attributes <name>                     The whitelist of attributes to query.

                                          This is a comma-separated list. If

                                          no attributes are specified all

                                          objects will be queried.

--date-format <format>                  The date format to use for formatting

                                          the time field. See java.text.

                                          SimpleDateFormat for options.

--help                                  Print usage information.

--jmx-url <service-url>                 The url to connect to to poll JMX

                                          data. See Oracle javadoc for

                                          JMXServiceURL for details. (default:

                                          service:jmx:rmi:///jndi/rmi://:

                                          9999/jmxrmi)

--object-name <name>                    A JMX object name to use as a query.

                                          This can contain wild cards, and

                                          this option can be given multiple

                                          times to specify more than one

                                          query. If no objects are specified

                                          all objects will be queried.

--reporting-interval <Integer: ms>      Interval in MS with which to poll jmx

                                          stats. (default: 2000)

可以这么使用

[iteblog@www.iteblog.com /]$ bin/kafka-run-class.sh kafka.tools.JmxTool --jmx-url service:jmx:rmi:///jndi/rmi://www.iteblog.com:1099/jmxrmi

运行上面命令前提是在启动kafka集群的时候指定export JMX_PORT= ，这样才会开启JMX。然后就可以通过上面命令打印出Kafka所有的metrics信息。

Kafka数据迁移工具

　　这个工具主要有两个：kafka.tools.KafkaMigrationTool和kafka.tools.MirrorMaker。第一个主要是用于将Kafka 0.7上面的数据迁移到Kafka 0.8（https://cwiki.apache.org/confluence/display/KAFKA/Migrating+from+0.7+to+0.8）；而后者可以同步两个Kafka集群的数据（https://cwiki.apache.org/confluence/pages/viewpage.action?pageId=27846330）。都是从原端消费Messages，然后发布到目标端。

[iteblog@www.iteblog.com /]$ bin/kafka-run-class.sh kafka.tools.KafkaMigrationTool --kafka.07.jar kafka-0.7.19.jar --zkclient.01.jar zkclient-0.2.0.jar --num.producers 16 --consumer.config=sourceCluster2Consumer.config --producer.config=targetClusterProducer.config --whitelist=.*

[iteblog@www.iteblog.com /]$ bin/kafka-run-class.sh kafka.tools.MirrorMaker --consumer.config sourceCluster1Consumer.config --consumer.config sourceCluster2Consumer.config --num.streams 2 --producer.config targetClusterProducer.config --whitelist=".*"

日志重放工具

　　这个工具主要作用是从一个Kafka集群里面读取指定Topic的消息，并将这些消息发送到其他集群的指定topic中：

[iteblog@www.iteblog.com /]$ bin/kafka-replay-log-producer.sh

Missing required argument "[broker-list]"

Option                                  Description

------                                  -----------

--broker-list <hostname:port>           REQUIRED: the broker list must be

                                          specified.

--inputtopic <input-topic>              REQUIRED: The topic to consume from.

--messages <Integer: count>             The number of messages to send.

                                          (default: -1)

--outputtopic <output-topic>            REQUIRED: The topic to produce to

--property <producer properties>        A mechanism to pass properties in the

                                          form key=value to the producer. This

                                          allows the user to override producer

                                          properties that are not exposed by

                                          the existing command line arguments

--reporting-interval <Integer: size>    Interval at which to print progress

                                          info. (default: 5000)

--sync                                  If set message send requests to the

                                          brokers are synchronously, one at a

                                          time as they arrive.

--threads <Integer: threads>            Number of sending threads. (default: 1)

--zookeeper <zookeeper url>             REQUIRED: The connection string for

                                          the zookeeper connection in the form

                                          host:port. Multiple URLS can be

                                          given to allow fail-over. (default:

                                          127.0.0.1:2181)

Simple Consume脚本

　　kafka-simple-consumer-shell.sh工具主要是使用Simple Consumer API从指定Topic的分区读取数据并打印在终端：

bin/kafka-simple-consumer-shell.sh --broker-list www.iteblog.com:9092 --topic test --partition 0

更新Zookeeper中的偏移量

　　kafka.tools.UpdateOffsetsInZK工具可以更新Zookeeper中指定Topic所有分区的偏移量，可以指定成 earliest或者latest：

[iteblog@www.iteblog.com /]$ bin/kafka-run-class.sh kafka.tools.UpdateOffsetsInZK

USAGE: kafka.tools.UpdateOffsetsInZK$ [earliest | latest] consumer.properties topic

需要指定是更新成earliest或者latest，consumer.properties文件的路径以及topic的名称