DSXchange

Your "option" has been given. Your design needs to change. Refer to my previous answer to which Ray agrees.

Actually the lock is only taken if the lookup fails. The hashed file is empty on the first run, because the table loading into the hashed file is empty. No rows will ever be found, so every row will obtain a lock. The job will degrade as it runs. Then, on the second run, the target table is full, a...

Make Vector is your best option at this point. When you perfect it, please post the solution for others.

Make sure you're not using locking under the reference lookup choices on the hashed file. You're probably filling the lock table with rows that never write back to the hashed file which releases the lock.

When a DS Designer connects to the repository, it gets a USERNO assigned. All jobs and routines are locked under the same USERNO value, so Ray's warning is to be careful to only release locks for a given job. But, if you trashed your network connection, all jobs under that DS Designer session would ...

Learn about job parameters, use your DS Designer manual for more information. They are the mechanism for feeding dynamic runtime control values to processes from the command line or the controlling processes (Batch, Sequence, custom job control).

There is no 10g stage. Continue with the 9i stage just keep using the 9i Oracle 32bit client and you're fine.

How about a job parameter? Such as

UNLOCK USER yournumber ALL is the best method for releasing locks. If you can, get your unix admin to go into the dsengine/bin directory and copy program list_readu to another file called "unlock". Now, you can execute list_readu from the unix prompt, and run unlock as well without going "into" the...

Not right now. I'm still working on developing a Bag of Tricks for PX, which could include such functionality. But it's not ready to share. Unless you have obscene volumes, you could still use a Server job prep the data using a hash file, and then dump it out to a sequential file. Then, use a PX job...

Efficient for speed? How are you going to control commit rate or rollback usage? Your design means that only 1 row of data is required for processing in order to fire this query. Which rows are affected? How many rows were actually updated? Why is it taking so long? How will you measure this job? If...

We do not have MetaStage, so is there no way I can get the job documentation or use the reporting assistant? We are on DS 7.5.1.A EE. Quite clearly answered by Ray. You will have to either purchase MetaStage or directly query the internal proprietary and undocumented internal repository and formula...

Your script can't possibly work, it's syntax is wrong. To feed commands to FTP you'll have to use indirection with a label. Example: ftp ${MACHINE} <<end_of_ftp>> ${LOGFILE} 2>&1 verbose ${FTPMODE} get ${SOURCE} ${TARGET} quit end_of_ftp RETVAL=$?

Make sure they don't try to use 64BIT 9.2 client with DS as only 32BIT 9.2 client is supported.

Whip out a standard shell script to run DB commands. Wrapper that script in a stage and use when needed.