DSXchange

Hi all, Thanks for your suggestions. My colleague developed the solution. First we add (sparcv9) to the beginning of 2nd record, store in a flat file, add a index column indicating CPUs belonging to Computer 1/2/3 etc. Second we read from the flat file with (sparcv9) as the delimiter. Rebuild the in...

Hi Craig,

Yes the last field is separated by newlines.

And yes its a typo. All the individual CPUs have a (sparcv9) at the end.

Thanks,
Jay

Hi All, Source is from a SQL Server. Target is an Oracle table. I have a requirement where the input is: Computerid=1,Name="Solaris",Value="CPU1 1156MHz (sparcv9) CPU2 1156Mhz . . CPU32 1156MHz" I have to ouput this to: SurrKey, Computerid=1,Name="Solaris",Value="C...

i have never worked with SQL Server... in Oracle i manually insert a row using TOAD taking a row from the flat file ... if you can insert one row you can insert pretty much all of them.... if problem persists, you may divide your file into smaller parts and get the problem rows...or get an idea whic...

most likely in some of the rows one of the columns is a null and DataStage is not able to figure it out...you may check the output to verify the result of your job...

There is an option in the transformer where in you can select "automatically map columns" having the same name. Other than that i dont know anything similar to the in.* you are asking about.

I think what Ray is telling is true. I had a similar problem in one of my jobs. I discussed the deadlock issue with my DBA and team members. We found that i was updating the same row in parallel processes. Once for Insert and once for Update. I think i had selected the Insert and Update option. This...

Hi all, I got the job to run. i have not done performance testing yet as i dont have a big enough test file. In the Oracle stage, in the query i say: nvl(to_number(substr(ZONE_ID,6,4),'9999'),'9999') as ZONE_CODE ZONE_ID is varchar(20). But actual value is like 1234.5678. Column ZONE_CODE is defined...

Yes. so performance is of vital importance.... also i will selecting from a load table and inserting into a partitioned table... already there are problems with sub-partitions...it seems there is a bug with Oracle 9i ver 9.2.0.4.0 if you create subpartitions, then the data dictionary gets corrupted....

thanks guys....i can see which one performs better....i'll keep u posted

my environment has not yet been set up....so might just take some time....

Thanks for the suggestions. Here are more details. The source will be an Oracle load table with a max of about 25 million records. The datatype is string, Varchar2(20), to be exact. Its basically 2 ids in 1 column. Target is also an Oracle table.But the columns all have their meaningful, respective,...

Thanks. I was not getting any ideas on modify stage either.

Hi All,

My requirement is this.

Split a column value , value1.value2,
to two separate columns containing value1 and value2.

e.x. Input col1=1234.5678 --> Output col1 = 1234 and Output col2=5678

What stage do i use other than the transformer?

DS ver = 7.5 EE(PX)

Thanks in advance
jay

Try to view 10 rows or 1 row at least...

i believe you shud check the Unix type termination in the sequential stage. your problem is because of incorrect line termination.