Thats because the column info:regioninfo has a serialized HRegionInfo.   You
need to take the cell bytes and deserialize them as a HRegionInfo instance
and then toString it.

See Writables in hbase utils.  It has a getHRegionInfo that takes bytes.

St.Ack

On Sat, Jul 25, 2009 at 9:23 AM, bharath vissapragada <
[email protected]> wrote:

> When i try to extract  rowResult.get("info:regioninfo").getValue() and
> convert it to string using Bytes.toString() and print it .. its printing
> junk characters ... whereas its working fine for other columngroups .. When
> i tried
> scan '.META.' from command line .. i observed that the value of
> info:regioninfo is not a string whereas for others it is a string!
>
> any comments?
>
> On Sat, Jul 25, 2009 at 9:38 PM, stack <[email protected]> wrote:
>
> > Just do a new Table(".META.") and scan it as you would any other table.
> >
> > Whats the error?
> >
> > St.Ack
> >
> > On Sat, Jul 25, 2009 at 3:03 AM, bharath vissapragada <
> > [email protected]> wrote:
> >
> > > Can you please tell me the classes using wch .META table can be scanned
> > ..
> > > I
> > > am getting error message when iam trying to scan it like a normal
> table.
> > >
> > > Thanks
> > >
> > > On Sat, Jul 25, 2009 at 3:14 PM, Ryan Rawson <[email protected]>
> wrote:
> > >
> > > > Nothing in the built in API does this for you, but you could scan
> > > > .META. looking for the regions of interest, then build scanners on
> the
> > > > specific ranges of each region.
> > > >
> > > > Good luck!
> > > >
> > > > On Sat, Jul 25, 2009 at 2:42 AM, bharath
> > > > vissapragada<[email protected]> wrote:
> > > > > Hi all ,
> > > > >
> > > > > Is there a way , i can get "Scanner" to the part of the table in a
> > > > specific
> > > > > region server using it's  "HOST NAME" ...
> > > > >
> > > > > Eg : suppose i have a table "A"  and one of the region servers have
> > > > HOSTNAME
> > > > > "region1"
> > > > >
> > > > > Can i get a scanner to all those rows of table "A" in "region1"..
> > > > >
> > > > > Thanks
> > > > >
> > > >
> > >
> >
>

Reply via email to