Flume三：负载均衡和故障转移（SinkProcessor）

案例一：故障转移

 agent1 : flume.conf（node2）FailoverSinkProcessor

#定义
a1.sources = r1
a1.channels = c1
a1.sinks = k1 k2
#定义sink组
a1.sinkgroups = g1

#source
a1.sources.r1.type = TAILDIR
# 用于断点续传，文件中包含各个路径下各个文件当前读取到的偏移量
a1.sources.r1.positionFile = /opt/flume_conf/tmp/tail_dir.json
# 文件组  如两目录 f1 f2
a1.sources.r1.filegroups = f1
# f1
a1.sources.r1.filegroups.f1 = /opt/data/.*.log
# 将数据流复制给所有 channel , 可以不用设置，默认是复制模式
a1.sources.r1.selector.type = replicating

#channel
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

#sink
# sink 端的 avro 是一个数据发送者
a1.sinks.k1.type = avro
a1.sinks.k1.hostname = node3
a1.sinks.k1.port = 4141

a1.sinks.k2.type = avro
a1.sinks.k2.hostname = node3
a1.sinks.k2.port = 4142

#sink的故障转移机制，agent2优先级高于agent3
a1.sinkgroups.g1.processor.type = failover
a1.sinkgroups.g1.processor.priority.k1 = 10
a1.sinkgroups.g1.processor.priority.k2 = 5
a1.sinkgroups.g1.processor.maxpenalty = 10000


#关联关系
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1
a1.sinks.k2.channel = c1
a1.sinkgroups.g1.sinks = k1 k2

 agent2 : flume.conf（node3）

#定义
a2.sources = r1
a2.sinks = k1
a2.channels = c1

#source
a2.sources.r1.type = avro
a2.sources.r1.bind = node3
a2.sources.r1.port = 4141

#channel
a2.channels.c1.type = memory
a2.channels.c1.capacity = 1000
a2.channels.c1.transactionCapacity = 100

#sink
#控制台日志打印
a2.sinks.k1.type = logger

#关联关系
a2.sources.r1.channels = c1
a2.sinks.k1.channel = c1

 agent3 : flume.conf（node3）

#定义
a3.sources = r1
a3.sinks = k1
a3.channels = c1

#source
a3.sources.r1.type = avro
a3.sources.r1.bind = node3
a3.sources.r1.port = 4142


# channel
a3.channels.c1.type = memory
a3.channels.c1.capacity = 1000
a3.channels.c1.transactionCapacity = 100

#sink 
#存储到hdfs
a3.sinks.k1.type = hdfs
a3.sinks.k1.hdfs.path=hdfs://mycluster/log_data/nginx/%Y%m%d/%H
#上传文件的前缀
a3.sinks.k1.hdfs.filePrefix = upload- 
#是否按照时间滚动文件夹
a3.sinks.k1.hdfs.round = true
#多少时间单位创建一个新的文件夹
a3.sinks.k1.hdfs.roundValue = 1
#重新定义时间单位
a3.sinks.k1.hdfs.roundUnit = hour
#是否使用本地时间戳
a3.sinks.k1.hdfs.useLocalTimeStamp = true
#积攒多少个 Event 才 flush 到 HDFS 一次
a3.sinks.k1.hdfs.batchSize = 100
#设置文件类型，可支持压缩
a3.sinks.k1.hdfs.fileType = DataStream
#多久生成一个新的文件
a3.sinks.k1.hdfs.rollInterval = 60
#设置每个文件的滚动大小大概是 128M
a3.sinks.k1.hdfs.rollSize = 134217700
#文件的滚动与 Event 数量无关
a3.sinks.k1.hdfs.rollCount = 0
#访问hdfs超时时间，5分钟
a1.sinks.k1.hdfs.callTimeout=300000

# 关联关系
a3.sources.r1.channels = c1
a3.sinks.k1.channel = c1

启动

依次执行

node3节点
flume-ng agent -n a3 -c conf -f /opt/flume_conf/flume3.conf
flume-ng agent -n a2 -c conf -f /opt/flume_conf/flume2.conf

node2节点
flume-ng agent -n a1 -c conf -f /opt/flume_conf/flume1.conf

测试：

            1、启动后查看 agent2 的logger控制台日志被打印出来，同时HDFS中没有新文件创建

            2、kill 掉 agent2 ，此时查看HDFS中已经有了对应的日志存储文件

案例二：负载均衡LoadBalancingSinkProcessor

 将案例一中的agent1配置中的sinkgroups的配置改为如下即可

a1.sinkgroups = g1
a1.sinkgroups.g1.sinks = k1 k2
a1.sinkgroups.g1.processor.type = load_balance
a1.sinkgroups.g1.processor.backoff = true
a1.sinkgroups.g1.processor.selector = random