DSXchange

MAXFETCHBUFF = 16384

Note the space either side of "=". These are necessary.
There is also MAXFETCHCOLS, for which the default is 400, meaning that your request can not return more than 400 columns.

AIX can support larger files than 2GB. Have your UNIX administrator adjust things. There's no need to upgrade DataStage; it won't help with this problem if AIX is still limited to 2GB files.

It's available directly in UniVerse 10.1 (via the Encrypt() and Decrypt() functions), but not in DataStage BASIC.

It's available directly in UniVerse 10.1 (via the Encode() function), but not in DataStage BASIC.

Show us the "job starting" entry for the job sequence - maybe it didn't have its parameters set.

See my recent post about MAXFETCHBUFF in uvodbc.config file.

You need a MAXFETCHBUFF setting in the uvodbc.config file. THe default maximum row size via ODBC is 8192 bytes.

Once it's a VarChar or Long VarChar you can use substrings quite happily. But does this really make sense for a BLOB? For a CLOB yes, but a BLOB?

You're 100% correct. An audit would need to show that the compilers were the same before you could assert that the jobs are the same.

Find out why they're getting locked, and avoid that behaviour. Don't kill the processes. Shut down clients gracefully. Learn how to determine who's got what locked.

Madhav_M wrote:Ray,
I used this option to generate log details.
Is there is any way to avoid manual intervention.
I mean can we execute any commands to automate this for all jobs in a folder?

Thanks
Madhav

No. There isn't.

Row Splitter can do the breakout into separate columns.

The only obvious this is that hmunx121 is not an IP address, and you don't have an explicit entry for [SRCDB] (or whatever else you want to call your DSN).

You still have to read the sequential file once to create the File Set or Data Set, so you process the entire volume of data twice.
If the sequential file has fixed width records you can set it up to be read in parallel.

Did you try to run the job from Director or from dsjob? If dsjob, make sure that the project name is correctly cased as well as correctly spelled. Try running from Director (where you simply can't get the project name wrong!).