DSXchange

Again, thank you for your reply. It seems odd that I can run the process and it works beautifully other than the joins passing the data in a "pipeline" manner. That is why I first assumed that as I watch it on the palette that it wasn't truly indicative of what is happening in the backgrou...

Arndw,

It appears that you had this same problem in 2004 but I don't see a solution in the thread.... I assume you found the problem or a work around?

viewtopic.php?t=96995&highlight=PASSWORD+MISSING

Well I played around with the environment variables and now I am getting the below messages when I have it set on DB2 partitioning and giving it nothing more than the table names in the options window for the partitioning. It acts like it needs credentials to do the partitioning but there is not a w...

First, Arndw thank you for your reply. Yes my DB is partitioned. I tried using the DB2 partitioning method but it had an issue resolving the instance even though I explicitly put it in there when it didn't like pulling it from the environment. Each DB2 operator has the instance and DB name in them a...

I have a parallel job that takes three tables and joins(inner) 2 together then runs that output through a transformer and then joins(inner) that output to the remaining table and that product loads back into the database. This is all on DB2 tables from the same Database. Myself and my superiors are ...

I remember years ago I received a diagram that had all of the operators listed out with small descriptions and it explained their nature as far as parallelism. I think this would be beneficial for a few on my team. If any of you know where I could find it, it would be greatly appreciated. I may can ...

Well I utilized the send buffer size variable to no avail. I ramped it up incrementally with no change. I am now looking into the DB2 params to make sure there is nothing there holding me back. Any other ideas or suggestions for me to research? I have already ran a trace on the network and it is mor...

If I create them it will intuitively know to modify those parameters of the system in correlation with my actions? There is nothing else I need to do other than create them like any other ENV Variable and give them a value that is indicative of the allocated space and run it?

I know someone here has tuned their system..... I fully understand that every system is different but I can't find anything else to change.... I am researching the system side as well.

Greetings All, I am trying to speed up the throughput from my server hosting DS and the Database. I am executing a parallel job that reads in a few DB2 tables, joins them, and then loads a table back on the DB. I have two nodes that house the DB, Coordinator Box 1 and Box 2. I can watch the performa...

ray.wurlod wrote:exec alone? That seems odd to me. ...

Tell me about it. I don't know what is going on. I will probably open a PMR soon if I can't resolve it. I really would just like to know why since I can work around it.

Have you all seen this scenario before? It's odd that it works perfect (DS immediately releases the job) when it builds a log but when I place the # in and comment the logging portion out DS holds onto the job until it abends. I am at a loss here. I reverted it back to log and pointed the log to a N...

I assumed that commenting out something was rather standard. DS is just calling these .ksh scripts and with it uncommented (no #) it works beautifully but commented (with #) it holds the job even though the execution in DS has a & trailing it. Before: logfile=$TEMPLOG/${log_script%\.ksh}.$dt exe...

Nobody............

I do this exact thing on a daily basis. How is the job designed on the pallet? When I first designed mine with nothing but Execute Commands on the pallett it would not run them all concurrently. For me to get things to run concurrently I had to build a Sequence job for each individual piece (job exe...