Actually trying to look how we can write a mapreduce program.
On Thu, Jun 7, 2012 at 7:12 AM, Harsh J <ha...@cloudera.com> wrote: > Ashish - Note though that Twitter's Snowflake is only roughly > sequential in generation. > > Abhishek - Do you really require proper sequence IDs? Why not just > keep and work with timestamps? > > On Thu, Jun 7, 2012 at 6:57 PM, Ashish <paliwalash...@gmail.com> wrote: >> Can try snowflake https://github.com/twitter/snowflake >> >> On Thu, Jun 7, 2012 at 6:20 PM, abhishek dodda >> <abhishek.dod...@gmail.com> wrote: >>> Hi all, >>> >>> I have a scenario where all my tables are in DB2 and after dumping >>> them into HDFS. >>> A unique sequence id (or) unique identifier an extra column should be >>> added at beginning or at the end of the table can anyone help me on >>> this please. >>> >>> Thanks >>> >>> Regards >>> Abhi. >> >> >> >> -- >> thanks >> ashish >> >> Blog: http://www.ashishpaliwal.com/blog >> My Photo Galleries: http://www.pbase.com/ashishpaliwal > > > > -- > Harsh J
