Hi Damien,
I have seen your patch, and I think it makes totally sense to do the modifications you are proposing. But in my case, I found out that I wasn't initializing the GoraInputFormat properly. In some way I was missing to set up the following line: GoraMapReduceUtils.setIOSerializations(context.getConfiguration(), true); And besides I was forgetting about the core-site.xml configurations related to IOSerializations that Gora needs. This is in [1]. So after that, all my troubles went away (: Hope we can put your patch in, as IMHO makes totally for running multiple instances of Gora Jobs in the same cluster. Nice job by the way! (: Renato M. [1] http://svn.apache.org/repos/asf/gora/trunk/gora-core/src/test/conf/core-site.xml 2013/8/31 Damien Raude-Morvan <[email protected]> > Le samedi 31 août 2013 11:22:24 Renato Marroquín Mogrovejo a écrit : > > Hi devs, > > Hi Renato, > > > I am in the middle of a war here trying to figure out how to use > > GoraInputFormat from outside GoraLand i.e. GoraMapper and friends. But I > am > > getting this exception when using HBase as backend: > [...] > > Seems like we are not persisting the partitioned query properly, but I am > > unsure this is really a problem as we are able to successfully run other > > MapReduce jobs e.g. QueryCounter, Nutch's jobs, and others. > > Any ideas, pointers, or ANYTHING is highly appreciated (: > > Thanks in advance. > [...] > > I encountered a very similar problem as you can see in JIRA bug tracker : > https://issues.apache.org/jira/browse/GORA-270 > > May you can try my attached patch, rebuild gora and see how it works... > > Regards, > -- > Damien >
