DSXchange

All failing "at the same time" sounds like a disk space issue to me. Any kind of "bigger than 2GB" situation would corrupt the hashed file and get you "blink" errors, do you see anything like that? If so, the solution to that is to switch to 64bit hashed files and that ...

priyadarshikunal wrote:

chulett wrote:perhaps switch to a bulk load.

Bulk load in parallel?

Sure, why not? Worst case, run it on one node, it may still be faster. There's far too little information in the original post to get better answers, so hopefully they'll pony up what you've asked for and whatever else might be helpful.

I have no idea what you are trying to add to this conversation in your post, can you try again please?

You should use IConv/OConv to convert that external format to YYYY-MM-DD, then append a zero time and use a Timestamp data type. The generated SQL will include the appropriate TO_DATE() function to get it into your DATE field all nice and proper like.

You did not get much in the way of an answer because (honestly) the answer really should have been fairly obvious - no. Unless you were looking for an option or switch of some kind to change the output format and again the answer would be no.

Are you certain that in all that patching / killing that your DataStage server wasn't simply rebooted while your jobs were running?

Odd... don't recall ever seeing such a message and a search for it here returned no other matches.

How precisely are you generating this surrogate key? Are you using the stage of the same name with a state file? That would certainly be the typical answer.

Comes down to what precisely you mean by 'combine'. You can certainly read and process multiple XML files, as many as you like in the same run, but the combining part may be up to you. Really depends on what that means to you.

Since it is already in that external format, just has a few extra bits, as noted the simplest solution would be to substring.

I would think it would be useful to post some sample data if you could, including the expected output from that input. That and explain exactly what the problem is 'to read this BITMAP part' so people can provide a little more specific help.

... and the resolution was?

With so little to go on... perhaps switch to a bulk load.

So... since there's no "link sort" in Server, is this actually a PX question? Or are you comparing a PX "link sort" with the Server sort stage? :? In a purely PX world, the short answer would be "none" with the noted exception that the stage gives you explicit control o...

They were still actually running or just in a 'running state'? Could you find active processes for the jobs?