DSXchange

I'll echo what has been said above: there are many, many, many different naming standards that can be used for links and stages. The best ones are internally consistent, easy to read (on the canvas, in the log, and in reports), and forbid duplicate link names within a single job. I've seen a wide va...

I'd check the version with just the copy stage against the original version. Replacing a transformer with a copy (when the copy has no output links) won't cause any new errors. If you are not getting that error in the original, then something has to be different that is causing the error in the test...

Hi Rajesh, Your issue is that all the columns are set to Not Null. Thus, there will never be a null value - even when the outer join is missing one of the inputs. What you want to be checking for instead of Null is the default value for the column. The other way to get around that is to set your out...

With the definition of the input given, I hope the correct order is guaranteed on the input, otherwise the addresses could get a bit jumbled. On a second note, is there a reason for recommending just one node? It should work just fine as long as your partitioning (I would probably use Hash partition...

I don't know the structure of your business-day table, but if it has a surrogate key - and you can have that key created as a monotonically increasing number by day, then your lookup becomes an easy one. Just grab the date with a SK that is 3 less than the SK for the input date. This can be faked in...

dsa,

Those options are each a different way to perform the same task. You only need 1 of the 3, not all 3.

I am curious if this is an interview question or if something in particular has prompted this??? Looking at your other posts it is hard to tell since it looks like you have some knowledge of the tool, but with curious gaps. I am hoping that this means you are working at a licensed client and are bei...

How fast does the ODBC stage run if you connect it to a copy (and nothing else)?

Are you sure it is the transformer that is the problem?

What you are seeing is expected behavior. If you don't need to identify what source your data came from (source 1, source 2, or both) then you just use a transformer after the join. In the transformer, your key derivations become: chat_id: if IsNull(1chat_id) then 2chat_id else 1chat_id contact_id: ...

Would it be possible to switch your FTP mode to binary? It sounds like the problems are due to trying to send a non-ascii file as ascii. chowdary_kish, you would be better served starting your own thread if a search doesn't give you the answer you need. Your problem is not identical to the original ...

There are way too many factors that play into the performance of any job for a general question to receive a truly meaningful and helpful answer. The more information you can supply, the better. What the environment is, what version of R/3, which BAPI is being called, how many nodes, etc., etc., etc...

Still doable, although it complicates things a bit more. Any other missing bits of information or requirements? My first thought is to build the job as: Input->Transformer->Funnel->Switch->Datasets The transformer would have 10 outputs so it can handle any combination of A and B identifiers in a sin...

I'll be a little bit surprised if you get any significant numbers who say they don't use best practices. That said, best practices will cover a wide range of tasks - from development, to change management, to deployment, to administration and more. I would expect that best practices that most people...

Thanks Ray. That is useful. My original thinking was about working on a monstrosity (100+ stages and constrained not to split it into smaller jobs) only want a peek on a small subset of the links - maybe even just 1 or 2. A second scenario came into my head where this idea would also be useful: Deal...

One of the frustrations that I encounter when doing debugging/troubleshooting is adding Peeks onto existing links. Depending on what stages are on the end of a link that you want to add a peek to - especially if you make changes in a less than optimal order - it can require redoing a lot of work [an...