Joined: 26 May 2005 Posts: 178 Location: Copenhagen, Denmark
It is something to do with utilites. This comes into picture when the key words SORTDATA,SORTDEVT SYSDA,SORTNUM key words. These are a kind of performance things.
For ex SORTKEYS while during LAOD would eliminate the need of supplied SYSUTI & SORTOUT datasets. It saves lot of I/O since it passes the key thru buffer. In manual you can find the concepts of subtasks which will be processed parallely there by improving performance. You can assign all the subtasks to a single UTPINT or you can have separate for each subtasks. Basically it says which kind of SORT technology was used. I couldn't address this in detail sorry. I'm sure you can find a very good stuff on this in IBM manual.
Joined: 05 Feb 2007 Posts: 199 Location: Sri Lanka
This contains messages from DFSORT. DFSORT is invoked internally by many DB2 utilities. You can code the DD like any other DD in JCL and provide an output file for it or direct it to sysout.