DSXchange

IBM's course code is DX506 - if you follow that link you'll get details of the class. The next scheduled class is October 2007, but with enough demand they can be convinced to run one sooner.

The authors of the product are Americans. They don't use the European "," as a decimal placeholder - even though they can cope with it in data.

Try entering 0.99 instead of 0,99 for Mprob.

What do YOU think that error message is telling you?

In particular, what function generated it?

Does your format string include percent signs and delimiters ("%mm/%dd/%yyyy")? Where have you set the format string?

Check out the IBM web site for your country. There will be a link to Education, or you can search for DX434 (parallel) or DX334 (server) - these are the codes for the Essentials classes.

The log file has become corrupted, perhaps because it reached 2GB at some point, perhaps for some other reason. Execute the following command from the Administrator command window. CLEAR.FILE RT_LOG543 You may need to manually reinstate the automatic log purge settings for this particular job from t...

In the Administrator command window execute the following commands. Let us know the outcome. LIST.ITEM VOC WITH @ID CONV "MCU" LIKE "...UTILITYWARNINGTOLOG..." LIST DSU_BP.O WITH @ID CONV "MCU" LIKE "...UTILITYWARNINGTOLOG..." Be careful - the first command is...

Didn't you like the answer yesterday in the Metadata forum?

Or you could search the forum for how to do it properly, preserving the log purge settings. This question has been answered before.

You can obtain this knowledge by attending the IBM classes or my classes, and probably others. The architecture is different in different editions, and different at design time, compile time, and run time, and may be different depending on whether you are executing in a single server, a cluster of m...

1. Don't load the server so heavily. 2. Reset the job. You may get a message logged "from previous run...". Read it and see what diagnosis it offers. 3. Investigate the offending stage to see if there's any possibility you may have done something that might throw an exception, for example attempting...

If it's truly recursive, you need a self-join (also known as a "fish hook join"), and therefore you really only have the option to define that join on the primary key (therefore the Surrogate Key). Or are you planning to snowflake the recursion. In that case you don't really have a Type 2 design, as...

Is the fix simply to re-enter the Mprob figure in the stage design?

It seems that Write mode in Oracle Enterprise stage places sqlldr log files into the scratch disk for each partition. Has anyone yet created a utility for gathering these into somewhere else? Also on the scratch disk are .par files (connection strings, in clear!), .log.bad files, fifo files and .ctl...

Very important to have APT_ORACLE_NO_OPS disabled (set to 0 or unset) if you're not using an RAC environment.