DSXchange

Do a search on this forum to get more info on the unix 9 kill.
Maybe you are funning over your limit of resources or processes at the unix level. If different jobs are aborting with the same error then its not a DS problem.
Try decreasing the amount of jobs you are running in parallel.

Tried with what?
Not having an after job subroutine or clearing the log file?
If this is an initial load then yes, you need to clear the table.
Do a load replace from /dev/null.
Do a search on how to truncate DB2 table. You will get the syntax.

Remove the after job subroutine and run the job. It should finish in time. Also, you really need to purge the log files more frequently, if the log file has a lot of entries, the routine will take a lot of time to find the correct time and status. Thats where a lot of your time is being eaten up.

Heres a post that i found which will be helpful.

If you are going for a user defined SQL, you might as well throw in a case statement in the sql itself and do this manipulation there.

Again. If Field1 needs is the Main group and Field2 is the sub-group then the sorting should be in that order, regardless. Switching the sort order will mess up the data. If performance does become an issue, resort to the unix sort.

Multiply it by 1
An arithmatic manipulation on a character forces it to become an integer (implicit conversion).
Another way

SQL3116W The field value in row "F0-106933" and column "22" is missing, but the target column is not nullable. Ok. Thats your bottle neck. You need to build a file that will get loaded without any warnings. Warnings like these slow down the process. Your load start and end times are impressive. Eve...

Yes. You DBA will be able to increase the page size for you. As for invoking the utility via command prompt, do this. -Go inside you load stage and set 'Load Immediate' and 'Remove Intermediate Datafile' to No. -Run the job. -Go to the location that you have specified in the property 'Directory for ...

The bulk stage does nothing but to create the three files that i mentioned earlier. Those files just invoke DB2's load utility. All that DataStage is doing here is building the BAT file and firing the load utility. Thats why i asked you to try to load the data using command prompt to invoke the load...

dsenv file basically contains all the enivronment variables that the server will reference during its startup and other programs would reference from time to time, like scripts etc. More info on dsenv can be found in dsbooks under the pdf Install and upgrade guide. It even tells you what variables a...

Ok. I think its the number of bytes you are sending is the bottleneck. Also towards the end of the message file, it gives you the time when the load was finished, when the load utility starts rebuilding the indices and when it finishes. See where is the bulk of the time going. Also i would advise to...

Are you sure the job remained unsuccessful in establishing a connection with the database? Whenever a database is involved in a job and the job seems to hang, sometimes its due to an alive thread to the database. Try what kris has suggested, if that fails, then you need to get in touch with you dba ...

As Craig said, it really depends how you want them grouped. Does Field2 come under Field1 or viceversa. The sort order will be specific to your requirement.

You need to add the connection information in the ODBC.ini file and the DSN name that you want in the drop down list in the uvodbc.config file.