I wonder if JDBC driver over Hive could help you. If you legacy ETL job can talk to a jdbc driver, it is a slow way of writing to HDFS and I don't have any experience doing it, e.g.: http://doc.cloveretl.com/documentation/UserGuide/index.jsp?topic=/com.cloveretl.gui.docs/docs/hive-connection.html
On Tue, Oct 8, 2013 at 10:07 AM, Jitendra Yadav <[email protected]>wrote: > Hi Bertrand, > > Thanks for your reply. > > As per my understanding mentioned open source does not support > procedure language(PL) flexibility. Right? > > I was looking for some other alternatives so that we can migrate our > existing code rather then creating java UDF etc. So handling complex > ETL business logic is still very difficult on Hadoop in terms of > coding, QA and performance? > > > Regards > Jitendra > > On 10/8/13, Bertrand Dechoux <[email protected]> wrote: > > open source : Pig, Hive, Cascading ... > > other : Talend ... > > > > Is that the answer you are expecting or are you looking for something > more > > specific? > > > > Regards > > > > Bertrand > > > > > > On Tue, Oct 8, 2013 at 5:47 PM, Jitendra Yadav > > <[email protected]>wrote: > > > >> Hi All, > >> > >> We are planning to consolidate our 3 existing warehouse databases to > >> Hadoop cluster, In our testing phase we have designed the target > >> environment and transferred the data from source to target (not in > >> sync but almost completed ). These legacy systems were using > >> traditional ETL/replication mechanism like Golden gate, Loaders, > >> PL/SQL language etc., FYI we are using 80% PL/SQL code and SQL server > >> packages in current environment, however we have re-writing some of > >> ETL jobs through java and python MR but looking for some more and easy > >> alternatives. > >> > >> What is best approach we should follow to complete this process? while > >> suggesting please take effort and timing in consideration( if > >> possible). > >> > >> Please guide. > >> > >> Regards > >> Jitendra > >> > > >
