DSXchange

Did you try giving the same server name and the port number for that particular instance?

Sounds great. Thats what the WFF activity stage is there.

There you go, you wanted a number, you have a number now by Ray.

Either it really is a server job or the OP is using a basic transformer. The constraints window does'nt look like its a px transformer.

Leave it for now. Try testing again tomorrow. If you face the same issue. Run a job that accesses this routine. You should be getting the correct result everytime you run it in a job.

Doing a sql level join is a personal preference as i like to keep my job design as simple as possible with the least number of stages used. But for complex and huge queries you have to use hashed files. Plus not all joins are possible within a single sql query. Plus if your tables are in different s...

If your lookup in a database stage then dont make them keys.
If your lookup in against a hashed file which was built with 3 keys. Then a simple unchecking the key box will not cut it. You have to rebuild your hashed file using a single key.

I am not Kumar.
Why dont you perform the bench marks and tell us which one is better, faster and why?
I did not get your second point, what does the creation of dataset have to do with CDC's performance

Let me clear one point. If you build a hashed file with two keys defined and try to read it with one or three. It will give you garbage. Secondly I prefer to do sql select join. This is done in the source database. If i cannot do sql joins then yes, i unload the keys and the columns I need only into...

As I noted earlier, I counsel against the use of NOTFOUND with db stages. YMMV. And the constraint, while 'unusual' in my book, should work fine - unless the sql used in the stage has been monkeyed with. I didnt read your previous reply. I dont use direct lookups to tables at all. I use a hashed fi...

More like C. Dont forget the osh engine. So I would say C and osh.

Whatever you make key has to be hit upon fromt he source for the lookup to be successful. I dont agree with the constraint that you are using. You should be using NOT(reflink.NOTFOUND). This way each it will pass only if lookup is successful. Dont do equality checks on individual key columns. Plus y...

Take a million rows, perform your bench mark. Increase it with factors of 10 untill you read maybe 1/3 of your data. Keep noting the performance change. This will give you a fairly accurate approximation of how much time it will take you for the full run.Its hard to guess the time frame without know...

Ok if the PPID is 1 that means its a zombie process and yes they need to be cleared. I thought you said no warnings. I assumed that the job finishes.
Anywho, I have'nt come across something like this, especially dealing with XML's. Maybe someone here can throw enlighten us with some more info.

Which is the case in most places. I have always seen them as RO and hence my assumption.