概念理解:
消息讀取方面
kafka中的消息,如果被讀取後,可以被重複讀取;
如果是被group A的用戶讀取了,其他組的用戶group B組的用戶可以讀取;
如果被組裏user1讀取了,組內其他成員B,在從頭開始讀取時,可以讀取到該數據;
Consumer group
http://www.cnblogs.com/huxi2b/p/6223228.html
個人認爲,理解consumer group記住下面這三個特性就好了:
consumer group下可以有一個或多個consumer instance,consumer instance可以是一個進程,也可以是一個線程
group.id是一個字符串,唯一標識一個consumer group
consumer group下訂閱的topic下的每個分區只能分配給某個group下的一個consumer(當然該分區還可以被分配給其他group)
http://www.cnblogs.com/huxi2b/p/6061110.html
http://www.cnblogs.com/huxi2b/p/6223228.html
本文着重討論了一下新版本的consumer group的內部設計原理,特別是consumer group與coordinator之間的交互過程,希望對各位有所幫助。
每次group進行rebalance之後,generation號都會加1,表示group進入到了一個新的版本,如下圖所示: Generation 1時group有3個成員,隨後成員2退出組,coordinator觸發rebalance,consumer group進入Generation 2,之後成員4加入,再次觸發rebalance,group進入Generation 3.
group與coordinator共同使用它來完成group的rebalance。目前kafka提供了5個協議來處理與consumer group coordination相關的問題:
1、Heartbeat請求:consumer需要定期給coordinator發送心跳來表明自己還活着
2、LeaveGroup請求:主動告訴coordinator我要離開consumer group
3、SyncGroup請求:group leader把分配方案告訴組內所有成員
4、JoinGroup請求:成員請求加入組
5、DescribeGroup請求:顯示組的所有信息,包括成員信息,協議名稱,分配方案,訂閱信息等。通常該請求是給管理員使用
Coordinator在rebalance的時候主要用到了前面4種請求。
組成員崩潰(member failure)
可以說離開組是主動地發起rebalance;而崩潰則是被動地發起rebalance。okay,
難道每次變更,generation都是會加1?
有相同的generation時,就會觸發rebalance;
Kafka在zk上的存儲
slider拉起的kafka在zookeeper中的存儲:存儲在目錄/kafka/client下;
zk的目錄下:
./zkCli.sh -server dcp11
/kafka/client
[zk: dcp11(CONNECTED) 10] ls /kafka/client/kafkajiu0522
[consumers, cluster, config, controller, isr_change_notification, admin, brokers, controller_epoch]
ls /kafka/client/kafkajiu0522/brokers/topics ----該kafka中的topic信息;
ls /kafka/client/kafkajiu0522/admin/delete_topics ---該kafka中刪除的topic信息;
kafka_server_jaas.conf文件的配置內容:
KafkaServer {
com.sun.security.auth.module.Krb5LoginModule required
useKeyTab=true
storeKey=true
useTicketCache=false
serviceName=kafka
keyTab="/etc/security/keytabs/kafkadcp18.keytab"
principal="kafka/[email protected]";
};
KafkaClient {
com.sun.security.auth.module.Krb5LoginModule required
useTicketCache=true;
};
Client {
com.sun.security.auth.module.Krb5LoginModule required
useKeyTab=true
storeKey=true
useTicketCache=false
serviceName=zookeeper
keyTab="/etc/security/keytabs/kafkadcp18.keytab"
principal="kafka/[email protected]";
};
命令
關於kafka啓動停止、topic的增查刪
創建topic:
./bin/kafka-topics --zookeeper DCP185:2181,DCP186:2181,DCP187:2181/kafka --topic test2 --replication-factor 1 --partitions 2 --create
bin/kafka-create-topic.sh --replica 2 --partition 8 --topic test --zookeeper 192.168.197.170:2181,192.168.197.171:2181
創建名爲test的topic, 8個分區分別存放數據,數據備份總共2份
刪除topic:
/data/data1/confluent-3.0.0/bin/kafka-topics --zookeeper DCP185:2181,DCP186:2181,DCP187:2181/kafka --delete --topic topicdf02175
查看topic列表:
/data/data1/confluent-3.0.0/bin/kafka-topics --zookeeper DCP185:2181,DCP186:2181,DCP187:2181/kafka --list
./kafka-topics --zookeeper DCP187:2181/kafkakerberos --list
啓動
nohup sh kafka-server-start ../etc/kafka/server.properties &
( 先用kerberos用戶登錄到環境中:kinit -kt /root/kafka.keytab [email protected] );
停止:./kafka-server-stop ../etc/kafka/server.properties &
slider上創建topic命令:注意和kafkaserver的創建有點不同:
./kafka-topics --zookeeper DCP185:2181,DCP186:2181,DCP187:2181/kafka/client/kafka04061 --topic topic04102 --replication-factor 1 -partitions 1 -create
生產消費:
有kerberos認證的,需要加上producer、consumer的對應配置文件:
./kafka-console-producer --broker-list DCP187:9092 --topic test2 --producer.config ../etc/kafka/producer.properties
選擇用bootstrap方式消費的:
./kafka-console-consumer.sh --from-beginning --topic topic05221 --new-consumer --consumer.config ../config/consumer.properties --bootstrap-server dcp11:9092
選擇用zookeeper方式消費的:
./kafka-console-consumer --zookeeper DCP185:2181,DCP186:2181,DCP187:2181/kafka --from-beginning --topic test2 --new-consumer --consumer.config ./etc/kafka/consumer.properties --bootstrap-server DCP187:9092
kafka不同消費方式後,記錄存儲的數據文件一樣:
無kerberos認證的:
用zookeeper方式消費:./kafka-console-consumer.sh --zookeeper DCP185:2181,DCP186:2181,DCP187:2181/kafka/client/kafka04112 --from-beginning --topic topic1;
用bootstrap-server方式消費:./kafka-console-consumer.sh --bootstrap-server DCP186:39940 --topic topic1 --from-beginning, --bootstrap-servers必須是kafka集羣中的其他一臺;
存儲的數據文件是:topic1-1和__consumer_offsets-0
存儲文件,消費到數據後,每個分區存放一個文件目錄;
關於acl權限:設置、查看、刪除
給普通用戶設置讀寫權限:
給client用戶賦producer寫權限:
./kafka-acls --authorizer-properties zookeeper.connect=DCP185:2181,DCP186:2181,DCP187:2181/kafka --add --allow-principal User:client --producer --topic test1
給client用戶賦consumer讀權限:
./kafka-acls --authorizer-properties zookeeper.connect=ai185:2181,ai186:2181,ai187:2181/kafka1017 --add --allow-principal User:client --consumer --topic test --group test-consumer-group
刪除所有權限
./kafka-acls --authorizer-properties zookeeper.connect=dcp18:2181,dcp16:2181,dcp19:2181/kafkakerberos --remove --producer --topic topicout05054
查看acl
./kafka-acls --authorizer-properties zookeeper.connect=DCP185:2181,DCP186:2181,DCP187:2181/kafka --list --topic test1
kafka不區分hostname;
常見問題及解決方法
啓動報錯
第一種錯誤
2017-02-17 17:25:29,224] FATAL Fatal error during KafkaServer startup. Prepare to shutdown (kafka.server.KafkaServer)
kafka.common.KafkaException: Failed to acquire lock on file .lock in /var/log/kafka-logs. A Kafka instance in another process or thread is using this directory.
at kafka.log.LogManager$$anonfun$lockLogDirs$1.apply(LogManager.scala:100)
at kafka.log.LogManager$$anonfun$lockLogDirs$1.apply(LogManager.scala:97)
at scala.collection.TraversableLike$$anonfun$map$1.apply(TraversableLike.scala:234)
at scala.collection.TraversableLike$$anonfun$map$1.apply(TraversableLike.scala:234)
at scala.collection.IndexedSeqOptimized$class.foreach(IndexedSeqOptimized.scala:33)
at scala.collection.mutable.WrappedArray.foreach(WrappedArray.scala:35)
at scala.collection.TraversableLike$class.map(TraversableLike.scala:234)
at scala.collection.AbstractTraversable.map(Traversable.scala:104)
at kafka.log.LogManager.lockLogDirs(LogManager.scala:97)
at kafka.log.LogManager.<init>(LogManager.scala:59)
at kafka.server.KafkaServer.createLogManager(KafkaServer.scala:609)
at kafka.server.KafkaServer.startup(KafkaServer.scala:183)
at io.confluent.support.metrics.SupportedServerStartable.startup(SupportedServerStartable.java:100)
at io.confluent.support.metrics.SupportedKafka.main(SupportedKafka.java:49)
解決方法:Failed to acquire lock on file .lock in /var/log/kafka-logs.--問題原因是有其他的進程在使用kafka,ps -ef|grep kafka,殺掉使用該目錄的進程即可;
第二種錯誤:對index文件無權限
把文件的權限更改爲正確的用戶名和用戶組即可;
目錄/var/log/kafka-logs/,其中__consumer_offsets-29是偏移量;
第三種生產消費報錯:jaas連接有問題
kafka_client_jaas.conf文件配置有問題
16環境上
/opt/dataload/filesource_wangjuan/conf下kafka_client_jaas.conf
KafkaClient {
com.sun.security.auth.module.Krb5LoginModule required
useKeyTab=true
storeKey=true
keyTab="/home/client/keytabs/client.keytab"
serviceName="kafka"
principal="client/[email protected]";
};
生產報錯
第一種:生產者向topic發送消息失敗:
[2017-03-09 09:16:00,982] [ERROR] [startJob_Worker-10] [DCPKafkaProducer.java line:62] produceR向topicdf02211發送信息出現異常
org.apache.kafka.common.KafkaException: Failed to construct kafka producer
at org.apache.kafka.clients.producer.KafkaProducer.<init>(KafkaProducer.java:335)
原因是配置文件:kafka_client_jaas.conf中配置有問題,keyTab的路徑不對,導致的;
第二種:生產消費報錯: Failed to construct kafka producer
報錯關鍵信息:Failed to construct kafka producer
解決方法:配置文件問題:KafkaClient中serviceName應該是kafka,之前配置成了zookeeper;重啓後,就好了;
配置文件如下:
KafkaServer {
com.sun.security.auth.module.Krb5LoginModule required
useKeyTab=true
storeKey=true
useTicketCache=false
serviceName=kafka
keyTab="/etc/security/keytabs/kafka.service.keytab"
principal="kafka/[email protected]";
};
KafkaClient {
com.sun.security.auth.module.Krb5LoginModule required
useKeyTab=true
storeKey=true
serviceName=kafka
keyTab="/etc/security/keytabs/kafka.service.keytab"
principal="kafka/[email protected]";
};
Client {
com.sun.security.auth.module.Krb5LoginModule required
useKeyTab=true
storeKey=true
useTicketCache=false
serviceName=zookeeper
keyTab="/etc/security/keytabs/kafka.service.keytab"
principal="kafka/[email protected]";
};
問題描述:
[kafka@DCP16 bin]$ ./kafka-console-producer --broker-list DCP16:9092 --topic topicin050511 --producer.config ../etc/kafka/producer.properties
org.apache.kafka.common.KafkaException: Failed to construct kafka producer
at org.apache.kafka.clients.producer.KafkaProducer.<init>(KafkaProducer.java:335)
at org.apache.kafka.clients.producer.KafkaProducer.<init>(KafkaProducer.java:188)
at kafka.producer.NewShinyProducer.<init>(BaseProducer.scala:40)
at kafka.tools.ConsoleProducer$.main(ConsoleProducer.scala:45)
at kafka.tools.ConsoleProducer.main(ConsoleProducer.scala)
Caused by: org.apache.kafka.common.KafkaException: java.lang.IllegalArgumentException: Conflicting serviceName values found in JAAS and Kafka configs value in JAAS file zookeeper, value in Kafka config kafka
at org.apache.kafka.common.network.SaslChannelBuilder.configure(SaslChannelBuilder.java:86)
at org.apache.kafka.common.network.ChannelBuilders.create(ChannelBuilders.java:70)
at org.apache.kafka.clients.ClientUtils.createChannelBuilder(ClientUtils.java:83)
at org.apache.kafka.clients.producer.KafkaProducer.<init>(KafkaProducer.java:277)
... 4 more
Caused by: java.lang.IllegalArgumentException: Conflicting serviceName values found in JAAS and Kafka configs value in JAAS file zookeeper, value in Kafka config kafka
at org.apache.kafka.common.security.kerberos.KerberosLogin.getServiceName(KerberosLogin.java:305)
at org.apache.kafka.common.security.kerberos.KerberosLogin.configure(KerberosLogin.java:103)
at org.apache.kafka.common.security.authenticator.LoginManager.<init>(LoginManager.java:45)
at org.apache.kafka.common.security.authenticator.LoginManager.acquireLoginManager(LoginManager.java:68)
at org.apache.kafka.common.network.SaslChannelBuilder.configure(SaslChannelBuilder.java:78)
... 7 more
[kafka@DCP16 bin]$ ./kafka-console-producer --broker-list DCP16:9092 --topic topicin050511 --producer.config ../etc/kafka/producer.properties
消費時報錯: ERROR Unknown error when running consumer: (kafka.tools.ConsoleConsumer$)
[root@DCP16 bin]# ./kafka-console-consumer --zookeeper dcp18:2181,dcp16:2181,dcp19:2181/kafkakerberos --from-beginning --topic topicout050511 --new-consumer --consumer.config ../etc/kafka/consumer.properties --bootstrap-server DCP16:9092
[2017-05-07 22:24:37,479] ERROR Unknown error when running consumer: (kafka.tools.ConsoleConsumer$)
org.apache.kafka.common.KafkaException: Failed to construct kafka consumer
at org.apache.kafka.clients.consumer.KafkaConsumer.<init>(KafkaConsumer.java:702)
at org.apache.kafka.clients.consumer.KafkaConsumer.<init>(KafkaConsumer.java:587)
at org.apache.kafka.clients.consumer.KafkaConsumer.<init>(KafkaConsumer.java:569)
at kafka.consumer.NewShinyConsumer.<init>(BaseConsumer.scala:53)
at kafka.tools.ConsoleConsumer$.run(ConsoleConsumer.scala:64)
at kafka.tools.ConsoleConsumer$.main(ConsoleConsumer.scala:51)
at kafka.tools.ConsoleConsumer.main(ConsoleConsumer.scala)
Caused by: org.apache.kafka.common.KafkaException: javax.security.auth.login.LoginException: Could not login: the client is being asked for a password, but the Kafka client code does not currently support obtaining a password from the user. not available to garner authentication information from the user
at org.apache.kafka.common.network.SaslChannelBuilder.configure(SaslChannelBuilder.java:86)
at org.apache.kafka.common.network.ChannelBuilders.create(ChannelBuilders.java:70)
at org.apache.kafka.clients.ClientUtils.createChannelBuilder(ClientUtils.java:83)
at org.apache.kafka.clients.consumer.KafkaConsumer.<init>(KafkaConsumer.java:623)
... 6 more
Caused by: javax.security.auth.login.LoginException: Could not login: the client is being asked for a password, but the Kafka client code does not currently support obtaining a password from the user. not available to garner authentication information from the user
at com.sun.security.auth.module.Krb5LoginModule.promptForPass(Krb5LoginModule.java:899)
at com.sun.security.auth.module.Krb5LoginModule.attemptAuthentication(Krb5LoginModule.java:719)
at com.sun.security.auth.module.Krb5LoginModule.login(Krb5LoginModule.java:584)
at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:57)
at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
at java.lang.reflect.Method.invoke(Method.java:606)
at javax.security.auth.login.LoginContext.invoke(LoginContext.java:762)
at javax.security.auth.login.LoginContext.access$000(LoginContext.java:203)
at javax.security.auth.login.LoginContext$4.run(LoginContext.java:690)
at javax.security.auth.login.LoginContext$4.run(LoginContext.java:688)
at java.security.AccessController.doPrivileged(Native Method)
at javax.security.auth.login.LoginContext.invokePriv(LoginContext.java:687)
at javax.security.auth.login.LoginContext.login(LoginContext.java:595)
at org.apache.kafka.common.security.authenticator.AbstractLogin.login(AbstractLogin.java:69)
at org.apache.kafka.common.security.kerberos.KerberosLogin.login(KerberosLogin.java:110)
at org.apache.kafka.common.security.authenticator.LoginManager.<init>(LoginManager.java:46)
at org.apache.kafka.common.security.authenticator.LoginManager.acquireLoginManager(LoginManager.java:68)
at org.apache.kafka.common.network.SaslChannelBuilder.configure(SaslChannelBuilder.java:78)
衍生問題:
kafka生產消息就會報錯:
[2017-05-07 23:17:16,240] ERROR Error when sending message to topic topicin050511 with key: null, value: 0 bytes with error: (org.apache.kafka.clients.producer.internals.ErrorLoggingCallback)
org.apache.kafka.common.errors.TimeoutException: Failed to update metadata after 60000 ms.
把KafkaClient更改爲如下的配置,就可以 了:
KafkaClient {
com.sun.security.auth.module.Krb5LoginModule required
useTicketCache=true;
};
消費報錯
第一種錯誤:replication factor: 1 larger than available brokers: 0
消費時報錯:Error while executing topic command : replication factor: 1 larger than available brokers: 0
解決辦法:/confluent-3.0.0/bin 下重啓daemon
./kafka-server-stop -daemon ../etc/kafka/server.properties
./kafka-server-start -daemon ../etc/kafka/server.properties
然後zk重啓;sh zkCli.sh -server ai186;
/usr/hdp/2.4.2.0-258/zookeeper/bin/zkCli.sh --腳本的目錄
如果還報錯,可以查看配置文件中下面的配置:
zookeeper.connect=dcp18:2181/kafkakerberos; --是group名稱
第二種錯誤:TOPIC_AUTHORIZATION_FAILED
./bin/kafka-console-consumer --zookeeper DCP185:2181,DCP186:2181,DCP187:2181/kafka --from-beginning --topic wangjuan_topic1 --new-consumer --consumer.config ./etc/kafka/consumer.properties --bootstrap-server DCP187:9092
[2017-03-02 13:44:38,398] WARN The configuration zookeeper.connection.timeout.ms = 6000 was supplied but isn't a known config. (org.apache.kafka.clients.consumer.ConsumerConfig)
[2017-03-02 13:44:38,575] WARN Error while fetching metadata with correlation id 1 : {wangjuan_topic1=TOPIC_AUTHORIZATION_FAILED} (org.apache.kafka.clients.NetworkClient)
[2017-03-02 13:44:38,677] WARN Error while fetching metadata with correlation id 2 : {wangjuan_topic1=TOPIC_AUTHORIZATION_FAILED} (org.apache.kafka.clients.NetworkClient)
[2017-03-02 13:44:38,780] WARN Error while fetching metadata with correlation id 3 : {wangjuan_topic1=TOPIC_AUTHORIZATION_FAILED} (org.apache.kafka.clients.NetworkClient)
解決方法:配置文件中下面的參數中的User的U必須是大寫;
super.users=User:kafka
或者有可能是server.properties中的adver.listen的IP是不對的,有可能是代碼中寫死的IP;
第三種錯誤的可能的解決方法:
無法消費,則查看kafka的啓動日誌中的報錯信息:日誌文件的所屬組不對,應該是hadoop;
或者,查看kafka對應的zookeeper的配置後綴,是否已經更改,如果更改了,則topic需要重新生成纔行;
第三種錯誤:消費的tomcat報錯:
[2017-04-01 06:37:21,823] [INFO] [Thread-5] [AbstractCoordinator.java line:542] Marking the coordinator DCP187:9092 (id: 2147483647 rack: null) dead for group test-consumer-group
[2017-04-01 06:37:21,825] [WARN] [Thread-5] [ConsumerCoordinator.java line:476] Auto offset commit failed for group test-consumer-group: Commit offsets failed with retriable exception. You should retry committing offsets.
更改代碼中,tomcat的心跳超時時間如下:
沒有改之前的:;
./webapps/web/WEB-INF/classes/com/ai/bdx/dcp/hadoop/service/impl/DCPKafkaConsumer.class;
重啓後,日誌中顯示:
[2017-04-01 10:14:56,167] [INFO] [Thread-5] [AbstractCoordinator.java line:542] Marking the coordinator DCP187:9092 (id: 2147483647 rack: null) dead for group test-consumer-group
[2017-04-01 10:14:56,286] [INFO] [Thread-5] [AbstractCoordinator.java line:505] Discovered coordinator DCP187:9092 (id: 2147483647 rack: null) for group test-consumer-group.
創建topic時錯誤
創建topic時報錯:
[2017-04-10 10:32:23,776] WARN SASL configuration failed: javax.security.auth.login.LoginException: Checksum failed Will continue connection to Zookeeper server without SASL authentication, if Zookeeper server allows it. (org.apache.zookeeper.ClientCnxn)
Exception in thread "main" org.I0Itec.zkclient.exception.ZkAuthFailedException: Authentication failure
at org.I0Itec.zkclient.ZkClient.waitForKeeperState(ZkClient.java:946)
at org.I0Itec.zkclient.ZkClient.waitUntilConnected(ZkClient.java:923)
at org.I0Itec.zkclient.ZkClient.connect(ZkClient.java:1230)
at org.I0Itec.zkclient.ZkClient.<init>(ZkClient.java:156)
at org.I0Itec.zkclient.ZkClient.<init>(ZkClient.java:130)
at kafka.utils.ZkUtils$.createZkClientAndConnection(ZkUtils.scala:75)
at kafka.utils.ZkUtils$.apply(ZkUtils.scala:57)
at kafka.admin.TopicCommand$.main(TopicCommand.scala:54)
at kafka.admin.TopicCommand.main(TopicCommand.scala)
問題定位:是jaas文件有問題:
解決方法:server.properties文件中的super.user要和jaas文件中的keytab的principle一致;
server.properties:super.users=User:client
kafka_server_jaas.conf文件改爲:
KafkaServer {
com.sun.security.auth.module.Krb5LoginModule required
useKeyTab=true
storeKey=true
useTicketCache=false
serviceName=kafka
keyTab="/data/data1/confluent-3.0.0/kafka.keytab"
principal="[email protected]";
};
KafkaClient {
com.sun.security.auth.module.Krb5LoginModule required
useKeyTab=true
storeKey=true
keyTab="/home/client/client.keytab"
principal="client/[email protected]";
};
Client {
com.sun.security.auth.module.Krb5LoginModule required
useKeyTab=true
storeKey=true
useTicketCache=false
serviceName=zookeeper
keyTab="/home/client/client.keytab"
principal="client/[email protected]";
};
kafka的配置文件:appConfig.json和resources.json配置項說明
兩個文件的作用:
appConfig.json:可以覆蓋在 metainfo.json中定義的配置值;當你需要提供新的運行時間方差時,這些可能會被用到;
resources.json:規定屬於應用的每個部件類型所需要的yarn資源;
兩個文件的配置項含義,下面的3個配置項沒有查詢到,其他的都已經確定;
appConfig.json文件中的:
"site.global.app_user": "${USER_NAME}",
----?應用中使用的用戶?我們環境中是client --官網和百度都沒有找到其意義
"site.broker.instance.name": "${USER}/${CLUSTER_NAME}",
---broker的實例名稱??官網上和官網都沒有;72上沒有配置具體值
"site.server.port": "${KAFKA_BROKER.ALLOCATED_PORT}{PER_CONTAINER}",
---KAFKA_BROKER.ALLOCATED_PORT}{PER_CONTAINER---官網上和百度都沒有找到;
appConfig.json和resources.json各個配置項的含義分別如下:
appConfig.json文件的各項配置:
appConfig.json
{
"components": {
"broker": {},
"slider-appmaster": {
"slider.hdfs.keytab.dir":
"/user/client/.slider/keytabs/client", ----keytab的目錄;固定的,如果不存在,新建下即可;
"slider.am.login.keytab.name": "client.keytab",
---keytab文件,該文件必須關聯所有kafka server和slider server的principle;
"slider.keytab.principal.name": "client/[email protected]"
---本機上,該keytab對應的principle
}
},
"global": {
"site.server.log.segment.bytes": "1073741824",
---- 一個log文件最大值,如果超過則切換到新的log文件
"system_configs": "broker",
----slider的變量:發送給容器的配置類型列表,例如core-site,hdfs-site,hbase-site;
"site.global.app_user": "${USER_NAME}",
----應用中使用的用戶?不知道在哪裏配置的
"site.global.kafka_version": "kafka_3.0.0",
---kafka的版本:confluent-3.0.0
"site.broker.instance.name": "${USER}/${CLUSTER_NAME}",
---實例名稱??官網上沒有
"site.server.port":
"${KAFKA_BROKER.ALLOCATED_PORT}{PER_CONTAINER}", ---?沒有找到
"site.global.pid_file": "${AGENT_WORK_ROOT}/app/run/koya.pid",
---容器的進程id:ps -ef |grep containername
"site.server.num.network.threads": "3",
---處理網絡請求的線程數
"site.server.log.retention.check.interval.ms": "300000",
--日誌保持檢查間隔時間,默認300000ms ,5分鐘
"site.broker.xms_val": "819m",
--??下面網址中沒有找到:http://slider.incubator.apache.org/docs/slider_specs/application_instance_configuration.html#variable-naming-convention
"site.server.delete.topic.enable": "true",
---能夠刪除kafka開關,如果設置爲false,則刪除失敗;
"java_home": "/usr/jdk64/jdk1.7.0_67",
---slider的變量:java的家目錄,和環境中的一致
"site.server.num.recovery.threads.per.data.dir": "1",
---每個用來日誌恢復的數據目錄的線程數---需要更加精簡化
"site.server.log.dirs":
"/data/mfsdata/kafkashared/kafka022701/${@//site/global/app_container_tag}",
---數據存儲的目錄:kafka022701是命名隨意,而且每個服務的該目錄是唯一的,不可以重複;
"site.server.num.partitions": "1",
---每個topic的默認分區數,默認值爲1,需要更多時,可以改大該值;
"site.server.num.io.threads": "8",
----做磁盤I/O操作的線程數
"site.broker.zookeeper": "dcp187:2181,dcp186:2181,dcp185:2181",
---broker所在的zookeeper集羣
"site.server.log.retention.hours": "168",
----每個日誌最短保存時間,超過可刪除
"site.server.socket.request.max.bytes": "104857600",
---一個socket可以接受的最大請求大小
"site.global.app_install_dir": "${AGENT_WORK_ROOT}/app/install",
---應用安裝的家目錄:是應用的根目錄;其中AGENT_WORK_ROOT是:container_work_dirs,即容器的工作目錄;
"site.global.app_root":
"${AGENT_WORK_ROOT}/app/install/kafka_2.11-0.10.1.1",
--是應用的根目錄;其中AGENT_WORK_ROOT是:container_work_dirs,即容器的工作目錄;
"site.server.socket.send.buffer.bytes": "102400",
---socket 服務器使用的發送信息緩衝區
"site.server.socket.receive.buffer.bytes": "102400",
---socket 服務器使用的接收信息緩衝區
"application.def":
".slider/package/KOYA/slider-kafka-app-package-0.91.0-incubating.zip"
---slider的變量:應用程序定義包的位置,例如/slider/hbase_v096.zip
"create.default.zookeeper.node"
-----可選配置項:應用是否需要默認的zk節點;我們配置文件中缺少該配置,該值是什麼不清楚;yes/no?還是true/false?後續待驗證;---
},
"metadata": {},
"schema": "http://example.org/specification/v2.0.0"
}
resources.json文件的各項配置項:
{
"components": {
"slider-appmaster": {},
"KAFKA_BROKER": {
"yarn.memory": "1024",
--組件實例所需要的內存數量,單位爲MB;該值必須比分配給任何JVM的堆(heap)
大小大;因爲一個jvm會比單獨使用一個heap使用更多的內存;
"yarn.role.priority": "1",
--該組件的唯一的優先級:它爲獨立組件類型提供唯一的索引;
"yarn.container.failure.threshold": "10",
---在一個失敗窗口中,一個組件可以失敗的次數;如果設置爲0,則爲不可以失敗;
"yarn.vcores": "1", --所需要的虛擬核心的數量:vcore:virtual
core;
"yarn.component.instances": "3", ---
"yarn.container.failure.window.hours": "1"
--容器失敗的窗口,默認值爲6;如果窗口大小有更改,該值則必須明確的配置下;如果失敗超過了窗口,則失敗次數需要重置;
}
},
"global": {},
"metadata": {},
"schema": "http://example.org/specification/v2.0.0"
}