flume hdfs一些简单配置记忆

2017/02/19 11:03
阅读数 22

  1 ############################################
  2 #  producer config
  3 ############################################
  4 #agent section
  5 producer.sources = s
  6 producer.channels = c c1 c2
  7 producer.sinks = r h es
  8
  9 #source section
10 producer.sources.s.type =exec
11 producer.sources.s.command = tail -f /usr/local/nginx/logs/test1.log
12 #producer.sources.s.type = spooldir
13 #producer.sources.s.spoolDir = /usr/local/nginx/logs/
14 #producer.sources.s.fileHeader = true
15
16 producer.sources.s.channels = c c1 c2
17
18 producer.sources.s.interceptors = i
19 #不支持忽略大小写
20 producer.sources.s.interceptors.i.regex = .*\.(css|js|jpg|jpeg|png|gif|ico).*
21 producer.sources.s.interceptors.i.type = org.apache.flume.interceptor.RegexFilteringInterceptor$Builder
22 #不包含
23 producer.sources.s.interceptors.i.excludeEvents = true
24
25 ############################################
26 #  hdfs config
27 ############################################
28 producer.channels.c.type = memory
29 #Timeout in seconds for adding or removing an event
30 producer.channels.c.keep-alive= 30
31 producer.channels.c.capacity = 10000
32 producer.channels.c.transactionCapacity = 10000
33 producer.channels.c.byteCapacityBufferPercentage = 20
34 producer.channels.c.byteCapacity = 800000
35
36 producer.sinks.r.channel = c
37
38 producer.sinks.r.type = avro
39 producer.sinks.r.hostname  = 127.0.0.1
42 producer.sinks.r.port = 10101
43 ############################################
44 #  hdfs config
45 ############################################
46 producer.channels.c1.type = memory
47 #Timeout in seconds for adding or removing an event
48 producer.channels.c1.keep-alive= 30
49 producer.channels.c1.capacity = 10000
50 producer.channels.c1.transactionCapacity = 10000
51 producer.channels.c1.byteCapacityBufferPercentage = 20
52 producer.channels.c1.byteCapacity = 800000
53
54 producer.sinks.h.channel = c1
55
56 producer.sinks.h.type = hdfs
57 #目录位置
58 producer.sinks.h.hdfs.path = hdfs://127.0.0.1/tmp/flume/%Y/%m/%d
59 #文件前缀
60 producer.sinks.h.hdfs.filePrefix=nginx-%Y-%m-%d-%H
61 producer.sinks.h.hdfs.fileType = DataStream
62 #时间类型必加,不然会报错
63 producer.sinks.h.hdfs.useLocalTimeStamp = true
64 producer.sinks.h.hdfs.writeFormat = Text
65 #hdfs创建多长时间新建文件,0不基于时间
66 #Number of seconds to wait before rolling current file (0 = never roll based on time interval)
67 producer.sinks.h.hdfs.rollInterval=0
68 hdfs多大时新建文件,0不基于文件大小
69 #File size to trigger roll, in bytes (0: never roll based on file size)
70 producer.sinks.h.hdfs.rollSize = 0
71 #hdfs有多少条消息时新建文件,0不基于消息个数
72 #Number of events written to file before it rolled (0 = never roll based on number of events)
73 producer.sinks.h.hdfs.rollCount = 0
74 #批量写入hdfs的个数
75 #number of events written to file before it is flushed to HDFS
76 producer.sinks.h.hdfs.batchSize=1000
77 #flume操作hdfs的线程数(包括新建,写入等)
78 #Number of threads per HDFS sink for HDFS IO ops (open, write, etc.)
79 producer.sinks.h.hdfs.threadsPoolSize=15
80 #操作hdfs超时时间
81 #Number of milliseconds allowed for HDFS operations, such as open, write, flush, close. This number should be increased if many HDFS timeout operations are occurring.
82 producer.sinks.h.hdfs.callTimeout=30000

 

hdfs.round

 

false

 

Should the timestamp be rounded down (if true, affects all time based escape sequences except %t)

 

hdfs.roundValue

 

1

 

Rounded down to the highest multiple of this (in the unit configured using hdfs.roundUnit), less than current time.

 

hdfs.roundUnit

 

second

 

The unit of the round down value - second, minute or hour.

 

83 ############################################
84 #  elasticsearch config
85 ############################################
86 producer.channels.c2.type = memory
87 #Timeout in seconds for adding or removing an event
88 producer.channels.c2.keep-alive= 30
89 producer.channels.c2.capacity = 10000
90 producer.channels.c2.transactionCapacity = 10000
91 producer.channels.c2.byteCapacityBufferPercentage = 20
92 producer.channels.c2.byteCapacity = 800000
93
94 producer.sinks.es.channel = c2
95
96 producer.sinks.es.type = org.apache.flume.sink.elasticsearch.ElasticSearchSink
97 producer.sinks.es.hostNames = 127.0.0.1:9300
98 #Name of the ElasticSearch cluster to connect to
99 producer.sinks.es.clusterName = sunxucool
100 #Number of events to be written per txn.
101 producer.sinks.es.batchSize = 1000
102 #The name of the index which the date will be appended to. Example ‘flume’ -> ‘flume-yyyy-MM-dd’
103 producer.sinks.es.indexName = flume_es
104 #The type to index the document to, defaults to ‘log’
105 producer.sinks.es.indexType = test
106 producer.sinks.es.serializer = org.apache.flume.sink.elasticsearch.ElasticSearchLogStashEventSerializer
 

展开阅读全文
打赏
0
0 收藏
分享
加载中
更多评论
打赏
0 评论
0 收藏
0
分享
返回顶部
顶部