BTW, maybe we should create an issue on jira for this problem if the Cassandra committee think it is necessary.
On Tue, May 18, 2010 at 5:30 PM, Jeff Zhang <zjf...@gmail.com> wrote: > reuse is not possible, but I think it won't hard for cassandra to > implement a ColumnFamilyOutputFormat. In my opinion, the > ColumnFamilyOutputFormat should be easier to implement than > ColumnFamilyInputFormat, because you don't need to think about split > in the implementation of OutputFormat. > > > > > > On Tue, May 18, 2010 at 5:22 PM, Ran Tavory <ran...@gmail.com> wrote: >> hbase - yes. But is that reusable for cassandra? >> >> On Tue, May 18, 2010 at 12:17 PM, Jeff Zhang <zjf...@gmail.com> wrote: >>> >>> I believe it is possible to write result back to cassandra. If I >>> remember correctly, HBase has both InputFormat and OutputFormat for >>> hadoop. >>> >>> >>> >>> >>> On Tue, May 18, 2010 at 5:08 PM, Ran Tavory <ran...@gmail.com> wrote: >>> > In the wordcount example the process reads from cassandra and the result >>> > is >>> > written to a local file at /tmp/word_count* >>> > Is it possible to read from cassandra and write the result back to >>> > cassandra >>> > to a specified cf/row/column? >>> > I see that there exists a ColumnFamilyInputFormat but not >>> > ColumnFamilyOutputFormat or something like that >>> > >>> > (in http://svn.apache.org/repos/asf/cassandra/trunk/src/java/org/apache/cassandra/hadoop/) >>> > My knowledge about hadoop and mr is pretty basic so maybe I'm missing >>> > something simple, lmk, thanks! >>> >>> >>> >>> -- >>> Best Regards >>> >>> Jeff Zhang >> >> > > > > -- > Best Regards > > Jeff Zhang > -- Best Regards Jeff Zhang
