DSXchange

Only taking 1.23% CPU indicates that CPU is not the bottleneck. However, dsapi_slave is an agent of your client process - it is not the process executing the job. Processes running jobs usually run uvsh. I suspect (lacking other processes) that pid 1267 is the one running your job. Even this process...

You can tune the amount of memory that the Aggregator stage uses (for a particular job) using the final option in the DS.TOOLS menu on the server. However, if you actually run out of memory, all bets are off - you can tune for 16GB but, on a system with only 4GB, this would be foolish in the extreme.

In short, your project is getting larger. T30 relates to the number of hashed files (including repository tables) open simultaneously; the first message relates to a job not being in a runnable state - maybe because it's aborted or was stopped, maybe because it's been changed but not recompiled.

You can read about them in the Installation and Upgrade Guide. However, all the questions are asked by install.sh before it invokes the setup script.

The multiple input links I had - transaction grouping on one link is silly. Enabling it I missed. But I can not see the control described in the help (which, incidentally, contains a description of what "on skip" does). The only control on the Transaction Grouping page is Rows per transaction - the ...

Welcome aboard. :D

This is one of the standard fields, a NYSIIS match on the first primary word. What is your output file? Have you done anything to construct its file description?

[2]

Why did it resolve the problem? What exactly was the problem?

Welcome aboard. :D

Have you corrected LD_LIBRARY_PATH in dsenv to include the directory in which dscapiop is located?

This is simply an alert. The upstream stage has its "preserve partitioning" flag (on Advanced tab in stage properties) set to "Propagate". Change it to "Clear".

Open the job properties of the job sequence. Check the two left-hand check boxes (or all four) on the General tab Compilation Options frame.

Click on Help while on that page to learn more. And/or search the forum for the terms you find on that page.

You must create the "config" file, which is really just a text file containing answers to the questions that install.sh will ask, one line at a time. Read through install.sh to figure out what questions will be asked.

There is no technical upper limit on the number of hashed files in a job. There is the practical limit of being able to maintain them. Every unit of work you add to a job necessarily lengthens its exection time, all else being equal.

Code -3 (DSJE.BADPARAM) means "bad parameter name" - that is, the parameter NAME that you are giving is not the name of a parameter in that job. Check both spelling and casing.