DSXchange

See if thispost helps.

Wow, to many nested quotes. Anywho... How is your data coming in. Its not hard to line them up. You can do it at the unix level using awk or in datastage. Line them up, get the percentage value to find out how many positions come under top 10% and get those many items. Its a few lines of code added ...

Yea but demand for a patch. Why use DRS stage when oracle enterprise stage is there for the very same reason.

ray.wurlod wrote:Which is exactly how Field() works.

Hmm, did'nt know that. But makes sense. So it is safe to say that Field() is equivalent to INDEX() with substring, in efficiency?

Why +2 and +3 is given in Ans. Can I pass a file of values as a argument to the Function? I am unable to understand the SQL of getCMD. Welcome Aboard pravin1581 In order to calculate median, I had to get the entries sorted as they might not be sorted while coming in. I did not know how to sort the ...

I had the same enthusiastic reaction. But no cigar from the OP

srini.dw wrote:2. select TO_DATE(TO_CHAR(TO_NUMBER(TO_CHAR(SYSDATE,'MM'))-1)||'/01/'||TO_CHAR(SYSDATE,'YYYY'),'MM
/DD/YYYY') from dual;

This query will break every time its run in January.

Simple ,yes ; efficient, I dont know about that. I belive, using INDEX() to get the position of "(" and use it in substring ([1,x]) where x will be the position returned by INDEX().

The biggest restriction, IMHO, is the fact that the job will operate in sequential. Synonymous to putting a shelby in a city where road speed limit is 25 mph.
Incorporate the functionality in a px shared container.

I did not realize this post was not to be answered to.

kumar_s wrote:Ok, does any of the existing JobSequence running in that server without error?

That was another question on my head. Is any sequence in this environment working at all? We are a few posts short of 5 pages and still shooting in the dark.

I think it is possible but you will see the key in red. I would rather do it in two transformers. Test it out and let us know.

No my friend, it will do simultaneous reads from all of your source hashed files and load to a single hashed file.
How did you fix your problem? What was the issue?

Lookup_32,0: Could not open table file "/ascential/datastage/70/d02/DataStage/Datasets/lookuptable.20070305.2w25xmd" for mmapping: Permission denied The error message tells you that you do not have permissions to that folder/file. Does your datastage id have access to that particular folder and file.

Build another sequence job from scratch that does the exact some thing. Delete this one. See if the problem persists.