diff options
Diffstat (limited to 'scalding/src/main/scala/sandcrawler/HBaseCountJob.scala')
-rw-r--r-- | scalding/src/main/scala/sandcrawler/HBaseCountJob.scala | 10 |
1 files changed, 6 insertions, 4 deletions
diff --git a/scalding/src/main/scala/sandcrawler/HBaseCountJob.scala b/scalding/src/main/scala/sandcrawler/HBaseCountJob.scala index 815478b..863c16f 100644 --- a/scalding/src/main/scala/sandcrawler/HBaseCountJob.scala +++ b/scalding/src/main/scala/sandcrawler/HBaseCountJob.scala @@ -13,7 +13,9 @@ class HBaseCountJob(args: Args, colSpec: String) extends JobBase(args) with HBas HBaseBuilder.parseColSpec(colSpec) val Col: String = colSpec.split(":")(1) - HBaseCountJob.getHBaseSource(colSpec) + HBaseCountJob.getHBaseSource(args("hbase-table"), + args("zookeeper-hosts"), + colSpec) .read .fromBytesWritable(Symbol(Col)) .debug @@ -22,9 +24,9 @@ class HBaseCountJob(args: Args, colSpec: String) extends JobBase(args) with HBas } object HBaseCountJob { - def getHBaseSource(colSpec: String) = HBaseBuilder.build( - "wbgrp-journal-extract-0-qa", // HBase Table Name - "mtrcs-zk1.us.archive.org:2181", // HBase Zookeeper server (to get runtime config info; can be array?) + def getHBaseSource(hbaseTable: String, zookeeperHosts: String, colSpec: String) = HBaseBuilder.build( + hbaseTable, // HBase Table Name + zookeeperHosts, // HBase Zookeeper server (to get runtime config info; can be array?) List(colSpec), SourceMode.SCAN_ALL) } |