DSXchange

What version of Windows?

You're right, Google turns up nothing on this name, not even on the usenet side. I'd suggest tracking down this executable on your system and checking its properties. Sometime you can get a clue from there as to what it is.

While there is nothing in DataStage to support this functionality natively, that's not to say it can't be done. Depending on your clustering software, you should be able to make it at least support "fail over" from node to node. I've done this in the past on Tru64 with shared EMC storage, but since ...

MaheshKumar Sugunaraj wrote:Message : Cannot get exclusive access to log for job <jobname>

Usually, the rest of this message is something like: "job may be being monitored". Is there any chance someone has a monitor window open for this job?

Ok, let's see... You didn't really specify but I'm guessing this is a job that is deleting and recreating a hash file each time it runs. You are explicitly pathing it, hence the mkdbfile command it is using. One quick option would be to simply 'Clear' the file rather than delete and recreate it each...

So, I'm assuming that in this case there is persistent data in the HF, so as long as we didn't synchronizs the HF, we can find differences ? I don't think you should be seeing any differences. As before, it would help to know what kind of differences you think you are seeing, then we stand a chance...

But if you are looking at performance perspective, my suggestion is to bring all data and use Merge/Join/Lookup Stage to get the expected data. Keeping in mind, of course, that these are PX only stages. The only one that is directly applicable to Server jobs is the "Lookup" stage but newbies need t...

What do you mean by "gave us different results"? Can you give a specific example? Is it with all hash files or just certain ones - and if certain ones, anything in common amongst the ones with issues? Are they anything other than Dynamic (Type 30) hash files? Created inside the Project or explicitly...

Do you mean a user-defined query can not directly look up another user-defined query? Hard to say. I'm not at all sure what "directly look up" means to you. You can have one "user-defined query" in your source stage bringing data into the job. You can then, in a Transformer, use a reference link to...

Then simply bring in all rows from one and do the lookup against the other. You can also bring all of one into a hash, then stream the other in and do a lookup against the hash.

Try Field(dslink,"?-",2) to get the second field.

chucksmith wrote:When it does not increment, what value is being returned.

That's the question. I wondered about the routine as well, but as you noted problems with it would return a 1 - which would reset the sequence rather than cause it to not increment.

Well, that helped. Glad to see you didn't change the gist of it in your original post.

Still don't see anything wrong with it, however. Might be the kind of problem that needs to be looked at in context - i.e. in your job as a whole. Unless someone else has some insight here?

Nope. It's a pretty straight-forward check and I don't see any problems with the logic. What worries me is this quote: It looks something like this: I've seen too many times where we've chased our tales because someone typed something in instead of using cut-and-paste and either didn't get it quite ...

To elaborate on what I said earlier... depending on how, exectly, your tnsnames.ora file was created, DataStage may or may not be able to parse the names out of it correctly. This, in my experience, happens when you manually edit the file or make it look like anything other than exactly what DataSta...

Parameterize everything that could change from environment to environment. At a minimum you should need three Job Parameters for Oracle: DSN, User ID and Password. It sounds like you should add a fourth - Owner or Schema Name. Then in the OCI stage, simply reference the Job Parameter in the sql. For...