What do you need Ninad? There are general notes on running hbase MR jobs in the mapred package: http://hadoop.apache.org/hbase/docs/r0.19.1/api/org/apache/hadoop/hbase/mapred/package-summary.html. Discusses CLASSPATH on your hadoop cluster.
Thereafter, running the BuildIndexTable job, after doing above suggested setup, looks like you'd do: ./bin/hadoop org.apache.hadoop.hbase.mapred.BuildIndexTable Will probably barf and tell you the command line options you are missing -- probably target directory for lucene indices, the table to index and the columns in the table to consider. No harm studying the source code in absence of better documentation. IIRC, this job will build as many lucene indices as there are reducers. Start with a small table and a single reducer. See how it goes. You'll have to copy the index out of hdfs to use it I'd imagine. St.Ack On Tue, Apr 21, 2009 at 11:12 PM, Ninad Raut <[email protected]>wrote: > Thanks JD. > > Can u help me with this one with a small code snippet, as an example > ?... There is not much dicussion on this .. vl require your help in > this. > > On 4/21/09, Jean-Daniel Cryans <[email protected]> wrote: > > Ninad, > > > > There is no index apart from the primary key so you are right, out of > > the box HBase doesn't directly provide such a facility. You can either > > index your table by hand using preferably a MapReduce job or use the > > BuildTableIndex provided in mapred. See > > > http://hadoop.apache.org/hbase/docs/r0.19.1/api/org/apache/hadoop/hbase/mapred/BuildTableIndex.html > > > > J-D > > > > On Tue, Apr 21, 2009 at 5:04 AM, Ninad Raut <[email protected]> > > wrote: > >> Suppose I have a Column Family called Status: which holds status > >> value. If I want to find how many values are there which have > >> Status=UNANALYSED I cannot do so using hbase shell.. is there any > >> other way??? > >> > >> On 4/21/09, Billy Pearson <[email protected]> wrote: > >>> there is multi ways to query hbase there's > >>> hbase shell > >>> thrift > >>> rest > >>> java api > >>> and I thank a few more. > >>> > >>> The easiest with out having to write code or anything would be hbase > >>> shell > >>> if just wanting to check manually if something is there or the value of > >>> it. > >>> > >>> Billy > >>> > >>> "Ninad Raut" <[email protected]> > >>> wrote in message > >>> news:[email protected]... > >>>> Hi, > >>>> > >>>> As there is no easy way to query HBase, can pig be used to query HBase > >>>> tables? > >>>> If so, can any one give me an example as to how to use it.... > >>>> > >>>> > >>>> Regards, > >>>> Ninad. > >>>> > >>> > >>> > >>> > >> > > >
