upgrade to hadoop-0.13?

Hi all,

As you know, hadoop-0.13 was recently released and it brings
some
impressive improvements over hadoop-0.12.x series. So the
obvious
question is: should we switch to hadoop-0.13?

I have tested nutch with hadoop-0.13 with all basic jobs
(inject,
generate, fetch, parse, updatedb, invertlinks, index, dedup)
and they
work fine.

-- 
Doğacan Güney

Doğacan Güney wrote:
> Hi all,
> 
> As you know, hadoop-0.13 was recently released and it
brings some
> impressive improvements over hadoop-0.12.x series. So
the obvious
> question is: should we switch to hadoop-0.13?
> 
> I have tested nutch with hadoop-0.13 with all basic
jobs (inject,
> generate, fetch, parse, updatedb, invertlinks, index,
dedup) and they
> work fine.
> 

We need to start implementing a different caching mechanism
for objects 
that we thus far cached in a Configuration instance.
Respective methods 
in Configuration are now deprecated, and will be removed in
Hadoop 0.14. 
  See HADOOP-1343 for more details.

This change will affect a lot of places in our code, so it
would be best 
to do it long before the next Nutch release.


-- 
Best regards,
Andrzej Bialecki     <><
  ___. ___ ___ ___ _ _   __________________________________
[__ || __|__/|__||/|  Information Retrieval, Semantic Web
___|||__||  |  ||  |  Embedded Unix, System Integration
http://www.sigram.com 
Contact: info at sigram dot com

On 6/18/07, Andrzej Bialecki <abgetopt.org> wrote:
> Doğacan Güney wrote:
> > Hi all,
> >
> > As you know, hadoop-0.13 was recently released and
it brings some
> > impressive improvements over hadoop-0.12.x series.
So the obvious
> > question is: should we switch to hadoop-0.13?
> >
> > I have tested nutch with hadoop-0.13 with all
basic jobs (inject,
> > generate, fetch, parse, updatedb, invertlinks,
index, dedup) and they
> > work fine.
> >
>
> We need to start implementing a different caching
mechanism for objects
> that we thus far cached in a Configuration instance.
Respective methods
> in Configuration are now deprecated, and will be
removed in Hadoop 0.14.
>   See HADOOP-1343 for more details.
>
> This change will affect a lot of places in our code, so
it would be best
> to do it long before the next Nutch release.

Opened NUTCH-501 for this. I also attached a (draft) patch
there.

>
>
> --
> Best regards,
> Andrzej Bialecki     <><
>   ___. ___ ___ ___ _ _  
__________________________________
> [__ || __|__/|__||/|  Information Retrieval, Semantic
Web
> ___|||__||  |  ||  |  Embedded Unix, System
Integration
> http://www.sigram.com 
Contact: info at sigram dot com
>
>


-- 
Doğacan Güney