DSXchange

I'm guessing it's one of those single runs with unlimited warnings kind of thing...

You corrupted the log file, probably by extending it past the 2GB limit. At this point about all you can do is clear it by going to the Administrator and issuing a CLEAR.FILE against RT_LOG1056. Rerun the job with a row limit or set it to abort after a small number of warnings so you can see (and ad...

Oracle updates aren't considered to be failure when they update nothing. Only failures like ones where a duplicate would be created, or a required field would be nulled are the kind that Oracle will 'reject'. You need to use an existing key hashed file to verify the update needs to be done and, assu...

Why would there be different types if they were all the same?

If you actually use the HFC, you'll see that each static type relates to a specific key pattern. Fill in the other variables and then switch through them to see which type they correspond to by looking at the generated command line.

For that you would need to make change to uvconfig file setting 1 as the value for the parameter 64BIT_FILES(for new file created only(not existing ones)). No, you don't - and it's generally a Bad Idea. That is also something that would come from reading the related postings here. Just create the o...

Right... so it's not about the number of records but the total size of the hashed file. And if you need to cross the 2GB Barrier, create the hashed file as a 64BIT hashed file. At that point, you are basically limited by the amount of disk space you have on hand.

Not positive, but I think it's still a chargable option. Also heard it is no longer being actively supported and will be phased out in the next release.

If you want to throw some money at the problem, there is always the dreaded Click Pack option from Ascential. It adds specific URL and weblog stages, from what I recall, plus the ability to use Perl in Server jobs.

The single quotes around the parameter, it will get upset without them. Dat's all.

Small correction as I'm pretty sure it will need to look like this: WHERE yourdatacolumn >= TO_DATE('#dateparameter#', 'YYYY-MM-DD HH24:MI:SS') And it may need to change slightly if you use an actual DATE parameter type as it allows no time portion: WHERE trunc(yourdatacolumn) >= TO_...

Well... basically, you just define them differently in each Project. Either that or have a common file all Projects leverage that have several entries for each connection: xxx-dev, xxx-test, etc.

Is that something you tried and had a problem with?

In other words, there's probably no reason to have a job design that is only good for the initial run and then has to be changed for any subsequent runs. Unless, perhaps, you are in a position where the 'initial' load volume is significantly larger than the daily incremental loads or needs to be han...

Or build a Sequencer job.

There is no such thing as a single oracle date format. The format used depends upon your database and environment settings, and thus there is no single answer for DataStage. Actually, for DataStage there is. Timestamp. As mentioned, the search function will turn up a ton of conversations we've alre...

Arnd has already mentioned the first thing I would suggest - bump up the Max Rows in virtual memory setting. How big are your rows? Any idea what the average record length is?