You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@hbase.apache.org by lohit <lo...@gmail.com> on 2011/09/28 19:06:37 UTC
Running MapReduce jobs against HBase
Hi,
As of 0.90.x if one has to use mapreduce jobs packed with hbase jar, they
will have to setup HADOOP_CLASSPATH using the method described here (hbase
claspath command)
http://hbase.apache.org/docs/current/api/org/apache/hadoop/hbase/mapreduce/package-summary.html#package_description
Is this common practice among users? Are there any alternative people have
thought about and could share their experience.
--
Have a Nice Day!
Lohit
Re: Running MapReduce jobs against HBase
Posted by Stack <st...@duboce.net>.
On Wed, Sep 28, 2011 at 11:03 AM, lohit <lo...@gmail.com> wrote:
> Thanks for the reply Stack. So, it looks like this is norm.
> Reason I asked is, the only config needed (I think) is ZooKeeper
> quorum/port. And did anyone put that in hadoop-site.xml and worked for them.
> (Although I agree it is ugly)
>
You mean build the conf dir into the job jar? That should work.
St.Ack
Re: Running MapReduce jobs against HBase
Posted by lohit <lo...@gmail.com>.
Thanks for the reply Stack. So, it looks like this is norm.
Reason I asked is, the only config needed (I think) is ZooKeeper
quorum/port. And did anyone put that in hadoop-site.xml and worked for them.
(Although I agree it is ugly)
2011/9/28 Stack <st...@duboce.net>
> What you looking for Lohit?
>
> You can do as doc describes or put the hbase classes in the CLASSPATH
> of running mapreduce out on cluster or just bundle all of the hbase
> dependencies up in your fat job jar.
> What would you like instead boss?
>
> St.Ack
>
> On Wed, Sep 28, 2011 at 10:06 AM, lohit <lo...@gmail.com>
> wrote:
> > Hi,
> >
> > As of 0.90.x if one has to use mapreduce jobs packed with hbase jar, they
> > will have to setup HADOOP_CLASSPATH using the method described here
> (hbase
> > claspath command)
> >
> http://hbase.apache.org/docs/current/api/org/apache/hadoop/hbase/mapreduce/package-summary.html#package_description
> >
> > Is this common practice among users? Are there any alternative people
> have
> > thought about and could share their experience.
> >
> > --
> > Have a Nice Day!
> > Lohit
> >
>
--
Have a Nice Day!
Lohit
Re: Running MapReduce jobs against HBase
Posted by Stack <st...@duboce.net>.
What you looking for Lohit?
You can do as doc describes or put the hbase classes in the CLASSPATH
of running mapreduce out on cluster or just bundle all of the hbase
dependencies up in your fat job jar.
What would you like instead boss?
St.Ack
On Wed, Sep 28, 2011 at 10:06 AM, lohit <lo...@gmail.com> wrote:
> Hi,
>
> As of 0.90.x if one has to use mapreduce jobs packed with hbase jar, they
> will have to setup HADOOP_CLASSPATH using the method described here (hbase
> claspath command)
> http://hbase.apache.org/docs/current/api/org/apache/hadoop/hbase/mapreduce/package-summary.html#package_description
>
> Is this common practice among users? Are there any alternative people have
> thought about and could share their experience.
>
> --
> Have a Nice Day!
> Lohit
>