Kafka在zookeeper中的存儲

時間 2019-11-11

原文原文鏈接

1、Kafka在zookeeper中存儲結構圖node

2、分析

2.1　topic註冊信息

/brokers/topics/[topic] :算法

存儲某個topic的partitions全部分配信息併發

[zk: localhost:2181(CONNECTED) 1] get /brokers/topics/topic2

Schema:
{
    "version": "版本編號目前固定爲數字1",
    "partitions": {
        "partitionId編號": [
            同步副本組brokerId列表
        ],
        "partitionId編號": [
            同步副本組brokerId列表
        ],
        .......
    }
}
Example:
{
"version": 1,
"partitions": {
"2": [1, 2, 3],
"1": [0, 1, 2],
"0": [3, 0, 1],
}
}

2.2　partition狀態信息

/brokers/topics/[topic]/partitions/[0...N] 其中[0..N]表示partition索引號負載均衡

/brokers/topics/[topic]/partitions/[partitionId]/statedom

Schema:
{
"controller_epoch": 表示kafka集羣中的中央控制器選舉次數,
"leader": 表示該partition選舉leader的brokerId,
"version": 版本編號默認爲1,
"leader_epoch": 該partition leader選舉次數,
"isr": [同步副本組brokerId列表]
}
 
Example:
{
"controller_epoch": 1,
"leader": 3,
"version": 1,
"leader_epoch": 0,
"isr": [3, 0, 1]
}

2.3　Broker註冊信息

/brokers/ids/[0...N] oop

每一個broker的配置文件中都須要指定一個數字類型的id(全局不可重複),此節點爲臨時znode(EPHEMERAL)ui

Schema:
{
"jmx_port": jmx端口號,
"timestamp": kafka broker初始啓動時的時間戳,
"host": 主機名或ip地址,
"version": 版本編號默認爲1,
"port": kafka broker的服務端端口號,由server.properties中參數port肯定
}
 
Example:
{
"jmx_port": -1,
"timestamp":"1525741823119"
"version": 1,
"host": "hadoop1",
"port": 9092
}

2.4　Controller epoch

/controller_epoch --> int (epoch) spa

此值爲一個數字,kafka集羣中第一個broker第一次啓動時爲1，之後只要集羣中center controller中央控制器所在broker變動或掛掉，就會從新選舉新的center controller，每次center controller變動controller_epoch值就會 + 1; 線程

2.5　Controller註冊信息

/controller -> int (broker id of the controller) 存儲center controller中央控制器所在kafka broker的信息orm

Schema:
{
"version": 版本編號默認爲1,
"brokerid": kafka集羣中broker惟一編號,
"timestamp": kafka broker中央控制器變動時的時間戳
}
 
Example:
{
"version": 1,
"brokerid": 0,
"timestamp": "1525741822769"
}

2.6　補充Consumer and Consumer group

a.每一個consumer客戶端被建立時,會向zookeeper註冊本身的信息;
b.此做用主要是爲了"負載均衡".
c.同一個Consumer Group中的Consumers，Kafka將相應Topic中的每一個消息只發送給其中一個Consumer。
d.Consumer Group中的每一個Consumer讀取Topic的一個或多個Partitions，而且是惟一的Consumer；
e.一個Consumer group的多個consumer的全部線程依次有序地消費一個topic的全部partitions,若是Consumer group中全部consumer總線程大於partitions數量，則會出現空閒狀況;

舉例說明：

kafka集羣中建立一個topic爲report-log 4 partitions 索引編號爲0,1,2,3

假若有目前有三個消費者node：注意-->一個consumer中一個消費線程能夠消費一個或多個partition.

若是每一個consumer建立一個consumer thread線程,各個node消費狀況以下，node1消費索引編號爲0,1分區，node2費索引編號爲2,node3費索引編號爲3

若是每一個consumer建立2個consumer thread線程，各個node消費狀況以下(是從consumer node前後啓動狀態來肯定的)，node1消費索引編號爲0,1分區；node2費索引編號爲2,3；node3爲空閒狀態

總結：
從以上可知，Consumer Group中各個consumer是根據前後啓動的順序有序消費一個topic的全部partitions的。

若是Consumer Group中全部consumer的總線程數大於partitions數量，則可能consumer thread或consumer會出現空閒狀態。

2.7　Consumer均衡算法

當一個group中,有consumer加入或者離開時,會觸發partitions均衡.均衡的最終目的,是提高topic的併發消費能力.
1) 假如topic1,具備以下partitions: P0,P1,P2,P3
2) 加入group中,有以下consumer: C0,C1
3) 首先根據partition索引號對partitions排序: P0,P1,P2,P3
4) 根據(consumer.id + '-'+ thread序號)排序: C0,C1
5) 計算倍數: M = [P0,P1,P2,P3].size / [C0,C1].size,本例值M=2(向上取整)
6) 而後依次分配partitions: C0 = [P0,P1],C1=[P2,P3],即Ci = [P(i * M),P((i + 1) * M -1)]

2.8　Consumer註冊信息

每一個consumer都有一個惟一的ID(consumerId能夠經過配置文件指定,也能夠由系統生成),此id用來標記消費者信息.

/consumers/[groupId]/ids/[consumerIdString]

是一個臨時的znode,此節點的值爲請看consumerIdString產生規則,即表示此consumer目前所消費的topic + partitions列表.

consumerId產生規則：

StringconsumerUuid = null;
if(config.consumerId!=null && config.consumerId)
consumerUuid = consumerId;
else {
String uuid = UUID.randomUUID()
consumerUuid = "%s-%d-%s".format(
InetAddress.getLocalHost.getHostName, System.currentTimeMillis,
uuid.getMostSignificantBits().toHexString.substring(0,8));

}
String consumerIdString = config.groupId + "_" + consumerUuid;

[zk: localhost:2181(CONNECTED) 11] get /consumers/console-consumer-2304/ids/console-consumer-2304_hadoop2-1525747915241-6b48ff32

Schema:
{
"version": 版本編號默認爲1,
"subscription": { //訂閱topic列表
"topic名稱": consumer中topic消費者線程數
},
"pattern": "static",
"timestamp": "consumer啓動時的時間戳"
}
 
Example:
{
"version": 1,
"subscription": {
"topic2": 1
},
"pattern": "white_list",
"timestamp": "1525747915336"
}

2.9　Consumer owner

/consumers/[groupId]/owners/[topic]/[partitionId] -> consumerIdString + threadId索引編號

a) 首先進行"Consumer Id註冊";

b) 而後在"Consumer id 註冊"節點下注冊一個watch用來監聽當前group中其餘consumer的"退出"和"加入";只要此znode path下節點列表變動,都會觸發此group下consumer的負載均衡.(好比一個consumer失效,那麼其餘consumer接管partitions).

c) 在"Broker id 註冊"節點下,註冊一個watch用來監聽broker的存活狀況;若是broker列表變動,將會觸發全部的groups下的consumer從新balance.

2.10　Consumer offset

/consumers/[groupId]/offsets/[topic]/[partitionId] -> long (offset)

用來跟蹤每一個consumer目前所消費的partition中最大的offset

此znode爲持久節點,能夠看出offset跟group_id有關,以代表當消費者組(consumer group)中一個消費者失效,

從新觸發balance,其餘consumer能夠繼續消費.

2.11　Re-assign partitions

/admin/reassign_partitions

{
   "fields":[
      {
         "name":"version",
         "type":"int",
         "doc":"version id"
      },
      {
         "name":"partitions",
         "type":{
            "type":"array",
            "items":{
               "fields":[
                  {
                     "name":"topic",
                     "type":"string",
                     "doc":"topic of the partition to be reassigned"
                  },
                  {
                     "name":"partition",
                     "type":"int",
                     "doc":"the partition to be reassigned"
                  },
                  {
                     "name":"replicas",
                     "type":"array",
                     "items":"int",
                     "doc":"a list of replica ids"
                  }
               ],
            }
            "doc":"an array of partitions to be reassigned to new replicas"
         }
      }
   ]
}
 
Example:
{
  "version": 1,
  "partitions":
     [
        {
            "topic": "Foo",
            "partition": 1,
            "replicas": [0, 1, 3]
        }
     ]            
}

2.12　Preferred replication election

/admin/preferred_replica_election

{
   "fields":[
      {
         "name":"version",
         "type":"int",
         "doc":"version id"
      },
      {
         "name":"partitions",
         "type":{
            "type":"array",
            "items":{
               "fields":[
                  {
                     "name":"topic",
                     "type":"string",
                     "doc":"topic of the partition for which preferred replica election should be triggered"
                  },
                  {
                     "name":"partition",
                     "type":"int",
                     "doc":"the partition for which preferred replica election should be triggered"
                  }
               ],
            }
            "doc":"an array of partitions for which preferred replica election should be triggered"
         }
      }
   ]
}
 
例子:
 
{
  "version": 1,
  "partitions":
     [
        {
            "topic": "Foo",
            "partition": 1         
        },
        {
            "topic": "Bar",
            "partition": 0         
        }
     ]            
}

2.13　刪除topics

/admin/delete_topics

Schema:
{ "fields":
    [ {"name": "version", "type": "int", "doc": "version id"},
      {"name": "topics",
       "type": { "type": "array", "items": "string", "doc": "an array of topics to be deleted"}
      } ]
}
 
例子:
{
  "version": 1,
  "topics": ["foo", "bar"]
}