DSXchange

Hi all, As part of an audit necessity, we're obliged to begin capture of when users log-on to and log-off from the Datastage engine. Whilst we are looking at the operating system side to capture log-on and log-off activities, I need to address the same within Datastage itself and it would additional...

Cheers Ray, so that action's restricted to an administrator account then. I'll go that route instead, thanks

Hi guys, Regarding the UV command 'release', which takes arguments 'file' and 'record', if I have a job design that is locked by another user, what should arguments 'file' and 'record' equate to? I've got the active record locks as shown: Device.... Inode.... Netnode Userno Lmode Pid Login Id Item-I...

Hi Sainath, Debugging was one of my first thoughts on it, but that's locked down in production. One of my other team is arranging move of the job and data to a dev environment where we can debug it, so that's still outstanding. When you say in parallel check the content of the hashed file, do you me...

Is your key order indeed CLASS_TYPE_CDE then CLASS_CDE everywhere in the job? Typically when a lookup like this fails it is due to some key issue. ... Arnd, I've retraced every transformer etc prior to the point where the hashed file is being looked-up and written to, to rearrange so the three fiel...

No worries on the typo, that just threw me, but my brain is also frazzled in any case

Yep, agree on the caching side... which is where I'm at really, but having tried all of the above, as per original post, still have the same issue

Since you seem to know you have some records where a previously generated ID should be removed, this should also mean your lookup in the monitor should be > 0. Is it? Hi Craig, thanks for chipping in too - the more the merrier :D Not sure where the "previously generated ID should be removed bi...

Hi Sainath, thanks for your reply too. Try having the file creation and clear as a seperate job. Having them within the reference job raises issues of delayed removed. Yep, fair point, it's not that way as is, but I have replicated this essentially by doing a delete and create ahead of running the j...

Arnd, Here's the lot: Type & Field......... Field. Field........ Conversion.. Column......... Output Depth & Name.......... Number Definition... Code........ Heading........ Format Assoc.. @ID D 0 HF_SIMPLE_MAP_C 10L S LASS_IDs CLASS_ID D 1 MD0 CLASS_ID 11R S CLASS_CDE I FIELD(@ID,@TM CLASS_...

Fair enough, still clutching at straws given all the attempts to rule cause in or out as per my original post. I'm not sure what you're asking on the "if I see column numbers" bit... whereabouts should I see that? Is this a reference to using the 'Position' attribute?

OK, I tried that, amending it on both the read and write - same result so far.

Hi Arnd, Thanks for the nippy reply :) The hashed file used as reference is compound-keyed. It has 3 columns, the second and third (in order) being the keys. In respect of column order, in the two spots where the hashed file is read and written, the metadata is identical, columns in the same order, ...

Hi guys, We have an issue in a job at the moment which uses the common practice of a hashed file read and write within the same job. A lookup is performed against this file to determine if a record exists in it already. Where this is not the case, a new ID is generated from a DB2 sequence and the re...

It's "DSParams"...

Look up the IsValid function and use this as the base of your logic to define whether a record is a reject or otherwise.