Link Collector and Link Partitioner Stages

Post questions here relative to DataStage Server Edition for such areas as Server job design, DS Basic, Routines, Job Sequences, etc.

Moderators: chulett, rschirm, roy

Post Reply
bapajju
Participant
Posts: 82
Joined: Wed Nov 19, 2003 11:58 pm

Link Collector and Link Partitioner Stages

Post by bapajju »

Hi,

Is there any limitation for the number of rows processed by a Link Collector or Link Partitioner stage. What I mean is is there any threshold limit in terms of number of records after which the efficiency of Link Collector and Link Partitioner stage goes down? Kindly let me know.

Thanks in advance
ray.wurlod
Participant
Posts: 54607
Joined: Wed Oct 23, 2002 10:52 pm
Location: Sydney, Australia
Contact:

Post by ray.wurlod »

None documented, and no reason why there should be. Rows are buffered (you can increase the buffer sizes), but the downstream process is draining the buffer ideally at least as fast as the upstream process is populating it. If your downstream processing can not go as fast as your upstream processing, then handshaking must occur when the buffer becomes full to tell the upstream process to stop sending - this is all managed internally and is not configurable. But it's not a function of the number of rows. Maybe you need to make your downstream processing go faster. Use statistics gathering on the active stages to identify hot spots.
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
bapajju
Participant
Posts: 82
Joined: Wed Nov 19, 2003 11:58 pm

Post by bapajju »

Thanks a lot Ray. The problem was really not in Link Collector. The problem was in the HASH lookup the Lookup was very slow.

Infact Link Collector works very well.

Thanks a ton for your suggestion
ray.wurlod wrote:None documented, and no reason why there should be. Rows are buffered (you can increase the buffer sizes), but the downstream process is draining the buffer ideally at least as fast as the upstream process is populating it. If your downstream processing can not go as fast as your upstream processing, then handshaking must occur when the buffer becomes full to tell the upstream process to stop sending - this is all managed internally and is not configurable. But it's not a function of the number of rows. Maybe you need to make your downstream processing go faster. Use statistics gathering on the active stages to identify hot spots.
Post Reply