利用ruby通过Thrift接口批量插入HBase

因为工作中使用fluentd采集日志，需要使用ruby通过thrift接口插入数据到hbase中，网上搜索了一下，发现很多都是插入单个数据，没办法只好参考Python语言写的批量插入修改了代码，终于搞定。

$:.unshift File.expand_path('gen-rb', File.dirname(__FILE__))
require 'thrift'
require File.expand_path('gen-rb/hbase', File.dirname(__FILE__))
transport = Thrift::BufferedTransport.new(Thrift::Socket.new("localhost", "9090"))
transport.open
client = Apache::Hadoop::Hbase::Thrift::Hbase::Client.new(Thrift::BinaryProtocol.new(transport))

## Get Table meta data

def newpass( len )
    chars = ("a".."z").to_a + ("A".."Z").to_a + ("0".."9").to_a
    newpass = ""
    1.upto(len) { |i| newpass << chars[rand(chars.size-1)] }
    return newpass
end
columnvalue = "v"*10
batchmutations = []
starttime=Time.now
1000.times do |x|

  mutations=[]
  rowkey ="RK_" << Time.now.strftime('%Y%m%d_'<< newpass(15))
    5.times do |i|
        mutations << Apache::Hadoop::Hbase::Thrift::Mutation.new(:column => "f1:" << i.to_s,:value => columnvalue)
    end
    batchmutations << Apache::Hadoop::Hbase::Thrift::BatchMutation.new(:row => rowkey,:mutations => mutations)
end
client.mutateRows("testdb2", batchmutations)