The number of sessions used by an Informatica TPT API Stream workflow should be able to be set in the definition of the Stream connection. I don't have direct experience in this area, but I have been told that it allows for setting min and max sessions.
Like many Stream/Tpump jobs, the client providing the data (the "source" in Informatica terminology) could be the bottleneck. In your case it may be that the source not supplying data fast enough to keep many of the sessions busy. You could try some iterations changing the number of sessions to see what maximizes the throughput.
For additional tips in this area, you could post your question on one of the Teradata Forums, and see how others using Informatica define the number of sessions for their TPT streams jobs.
Mark,
The number of sessions used by an Informatica TPT API Stream workflow should be able to be set in the definition of the Stream connection. I don't have direct experience in this area, but I have been told that it allows for setting min and max sessions.
Like many Stream/Tpump jobs, the client providing the data (the "source" in Informatica terminology) could be the bottleneck. In your case it may be that the source not supplying data fast enough to keep many of the sessions busy. You could try some iterations changing the number of sessions to see what maximizes the throughput.
For additional tips in this area, you could post your question on one of the Teradata Forums, and see how others using Informatica define the number of sessions for their TPT streams jobs.
Thanks, -Carrie