DSXchange

They're not the same. Check not only the metadata in DataStage but also the metadata in Oracle and remember that it's nullability that you're checking, not data type (though that has to be the same too).

By "[NUL]" do you mean those five characters, do you mean the ASCII NUL character (Char(0)), or do you mean something else (and, if so, what)?

You could change the system clock while the job is running. :lol: You really have not provided enough information to attempt cogent diagnosis - we have no idea what you are trying to do with those Transformers and Oracle connections, nor whether you're doing it efficiently. Nor even why you want to ...

There is no problem at all using consecutive Sort stages, I have one example that uses three. Could this be an issue of improper partitioning? I am guessing that you are using a two-node configuration and Auto partitioning. This is not correct to achieve what you want to achieve.

dsscholar wrote:From today only it happens.

What is different "today"? Note that "nothing" is not the correct answer.

Definitely "maybe". There are too many other factors. Disabling operator combination is primarily to address systems that are CPU bound (processors being asked to do too much work because two many operators have been combined into one). But it's fair to say that this could also lead to an ...

To challenge the data requirement document is the correct course of action.

Yes of course you can remove them - maybe Trim() maybe Convert(). But it's an incorrect requirement to start with. Educate the authors about not requesting things that are just plain wrong. It will help you in future!

What does "delimiter not found" suggest to you?

The internal representation of NULL is Char(128). The ASCII NUL character is Char(0). Your "strange character" is probably one of these.

If EXECUTE doesn't work, try CALL.

The ability to be combinable can be set in individual stages on the Advanced tab. Global operator combination can, as Craig notes, be set using the appropriate environment variable, which I always add to parallel jobs as a parameter.

Setting array size to 0 logically transfers all records at once, but the network still breaks that mass of data into packets so any gain would be insignificant.

Array size has nothing at all to do with execution mode or read mode.

No it doesn't, but it would be easy enough to create them. Certainly accumulate totals, sums and sums of squares etc. and then just plug those into the requisite formulae. I've done this with skewness and kurtosis in the past. Alternately you can construct Build stage(s) to do exactly what you requi...

The answer to your first question is either "none" (nodes are a logical construct, they don't occur "in UNIX") or "all of them" (the number of nodes is as defined in your parallel execution configuration file). Which ones are "working" (whatever that means) ca...

The size of the group latch table is GSEMNUM x GLTABSZ and the size of the record lock table is GSEMNUM x RLTABSZ. Usually, when one is increased, the other is decreased so that the total amount of memory remains (approximately) unchanged. By increasing just one you took the size of the memory image...