sybase parallel extract

jasper · Post by **jasper** » Wed Jun 22, 2005 3:16 am

When using the Sybase OC stage to read data there is the option to work parallel (stage-advanced tab, execution mode).
In the oracle stage there is something likewise, but there you have some control about how it will run parallel because it askes for a partition table. In sybase we don't seem to have any control about how it works.

Does anyone know something about the internal mechanism how the partitiong of data is done(on oracle it uses rowid's) and maybe how we can control it?

related question: if we would use this stage on a sybase IQ, what would be the effect (There is a seperate IQ-load stage, but not for reading).

kcbland · Post by **kcbland** » Wed Jun 14, 2006 9:45 am

Since no one responded to this post, I am interested in hearing about experiences with Sybase IQ working with PX. How is the partitioning done on a table? Is a parallel read with divide and conquer method being used on a "SELECT * FROM table" type query?

jasper · Post by **jasper** » Thu Jun 15, 2006 12:22 am

well, luckily for me this was a one-time only experience.
I didn't get the parallel read working on IQ. If you define more then one node, it just executes the full query on every node.

ray.wurlod · Post by **ray.wurlod** » Thu Jun 15, 2006 2:07 am

Was the Sybase IQ table partitioned? Search the forum for how DB2 users included the partitioning variables in the WHERE clause.