You are viewing a plain text version of this content. The canonical link for it is here.

Posted to user@hbase.apache.org by Ted Yu <yu...@gmail.com> on 2010/01/22 20:08:36 UTC

concurrency in exporting HBase contents

Hi,
Suppose during export there is ongoing write operation to HBase table I am
exporting, which snapshot does export use ?
Is there special action I should take ?

Thanks

Re: concurrency in exporting HBase contents

Posted by Ted Yu <yu...@gmail.com>.

I am using MapReduce for import/export.

Can you forward me some links about multi datacenter replication ?

Thanks

On Fri, Jan 22, 2010 at 4:17 PM, Jean-Daniel Cryans <jd...@apache.org>wrote:

> Which kind of export are you talking about? A MapReduce or a distcp?
>
> In any case, it is very probable that your import will miss some
> writes unless you block them.
>
> In 0.21 this will be a lot easier using multi datacenter replication
> along with the ability to replay logs from one cluster to another
> starting from a certain point in time.
>
> J-D
>
> On Fri, Jan 22, 2010 at 11:08 AM, Ted Yu <yu...@gmail.com> wrote:
> > Hi,
> > Suppose during export there is ongoing write operation to HBase table I
> am
> > exporting, which snapshot does export use ?
> > Is there special action I should take ?
> >
> > Thanks
> >
>

Re: concurrency in exporting HBase contents

Posted by Jean-Daniel Cryans <jd...@apache.org>.

Which kind of export are you talking about? A MapReduce or a distcp?

In any case, it is very probable that your import will miss some
writes unless you block them.

In 0.21 this will be a lot easier using multi datacenter replication
along with the ability to replay logs from one cluster to another
starting from a certain point in time.

J-D

On Fri, Jan 22, 2010 at 11:08 AM, Ted Yu <yu...@gmail.com> wrote:
> Hi,
> Suppose during export there is ongoing write operation to HBase table I am
> exporting, which snapshot does export use ?
> Is there special action I should take ?
>
> Thanks
>