DSXchange

Please post the job score. You will need to set APT_DUMP_SCORE to True to achieve this - easiest way is to add it as a job parameter. Please also post your job design. Otherwise we'd be wasting time guessing at possible causes. There are several, for example lookups that return multiple rows, joins ...

Probably your support provider is best to ask. This sounds like a bug. What (exact) version of DataStage are you using? Can you post exactly how you are setting up the SCD stage for the Type 2 column?

Search the forum for "UPDATE SDKSequences" to learn how to initialize a DataStage sequence.

SDKSequences is a table (hashed file) where the sequences are stored. It is keyed by sequence name (the name you supply as the function argument) and contains the next value to be used in that sequence.

Not cached AND set for "lock for updates" would be my suggestion.

What qualifier?

Please mark the thread as Resolved.

Quelle embarrasant!

But not "my" band, please! DSXchange is Dennis's baby, I've just been sharing some of my good fortune (of having worked for the vendor for some years) on DSXchange.

Ah, the "Microsoft solution" again. Note that OLE Automation errors more usually occur on the client. In the case of DataStage, the objects in question may be things like the Session object through which your connection to the server is maintained. Re-starting the server would have necessitated your...

A full and general solution was given in my earlier reply.

Premium membership is worth it. Less than 30c per day. 100% of this goes to the hosting and bandwidth costs incurred by DSXchange.

Are you referring to the aesthetics, or simply the appearance, of the buttons. Apparently they don't look very good in IE, but are somewhat better if you use some other browser such as FireFox, NetScape or Safari.

Does the Format tab nominate tab as the field delimiter character? Change it to 000 (or any other character not in the data). You can't have tabs in tab-delimited data unless all character strings are quoted. More generally you can't have "x" in "x"-delimited data unless all character strings are qu...

sachin1, where did you read in the specification from basu.ds that there is a need to delete anything? "Reject records" is not the same as "delete records". Rejecting records is simply a matter of having a constraint on a Transformer stage output link that does not allow records to pass if the count...

Wrong. That function belongs in an SQL query, not in a Transformer stage. In a Transformer stage simply take the leftmost ten characters, or the first space-delimited field, from the timestamp. This works because there are no data types in server jobs. Left(InLink.TheTimestamp,10) Field(...

There are several de-duplication techinques available. The easiest is probably to run a query against the table SELECT keycol,COUNT(*) FROM table GROUP BY keycol HAVING COUNT(*) > 1; If you now want to compare that against the contents of some CSV file, load the SQL result into a has...

Quite simply because you're not specifying it correctly. Take the time to write out a specification of what you need to derive and then develop the algorithm that will derive it. The Field() function does work properly, but only when specified propertly. In your case, look very carefully at the deli...

As well as documenting the hardware you also need to document the resource consumption of each job per operator per node under full load scenario (this should be a part of testing, ideally). Then you figure out how many jobs you'd like to run simultaneously and on how many nodes each, such that the ...