DSXchange

PX is real fast but depends on how many processors you have set your PX config file to use. It tends to take over a majority of the CPU utilization for those processors. If you are using PX with just 4 processors then you should design your batch so that relatively very few jobs run during the PX jo...

an case to do it outside of datastage as a sql script spooling to a file.

That way you can put a parallel hint
/*+ full(table_name) parallel(table_name, */
and force it to go in parallel and the dump to file is not slowed down by the server throughput, or if you have PX do it in PX.

You could try this- If you need to send an e-mail for every row that f1 is > f2 then you can use the below logic within the datastage job f3 = f1:',':f2 DSSendMail("From:x@abc.com\nTo:y@abc.com\nSubject:Hi\nBody:f3\n") Or you can write all records with f1 > f2 to a file and call DSSenMail ...

Have you tried to 'Cleanup Resources' from Director for the job. Normally that clears out all the processes. If the lock on the job is still not released then you can try the Clear Status from Director. This should return the job to compiled status. If above does not work make sure no phatom process...

The PX job reads from a partitioned Oracle table and then copies the data set using the copy stage after which there is a lookup stage followed by a aggregator stage and finally the oracle upsert stages. The job has run fine until Oct 31st when it aborted with the " Operator terminated abnormal...

Is your filename to check - Users.csv - a job parameter? The Execute Command stage doesn't parse them correctly. Yes, the log entry looks just peachy but the actual command executed includes the untranslated job parameter, which may be what's tripping you up. Me, I'd do this in a routine and use DS...

I am using the test -s filename command in the ExecuteCommand stage to check for a an empty file. A return code of 0 continues control to the next job else the sequence aborts. However this does not work. Always aborts with following message whether the file is empty or not. "JobSeqLoadAdExtrac...

FYI...
I think I have found a round about way of doing this. I am exporting the jobs in the project in XML format and writing a job to parse it to look for a property value of 'Yes' for the 'CREATETABLE' node.

I have a routine to search jobs for a string but it does not work for this case since the 'Create Table' is an option in datastage and searching for string 'create table' does not return anything. Is there any specific string I need to look for to identify jobs using this option. In the logfiles on ...

Hi I was moving a few jobs via version control. Midwat through I got disconnected to the server. Connected back and found I could open one of the jobs. Tried to delete the job but got an error that could not open RT_BP333 and RT_BP333.O. Also when I tried to import a bunch of jobs again hit some err...

Here is some unix code you could use or modify. *********************************** dsjob -logsum Proj Jobseq > TEMPFIL1 for AbortJob2 in `cat TEMPFIL1 | egrep "\(Aborted\)|\(Finished with warnings\)" | cut -f 4 -d " "` do dsjob -logsum Proj Jobseq > TEMPFIL3 done ***************...

We have had many such issues in our jobs but mostly involving multiplication or division where the decimal precision changes. So we use the round function within Oracle so that we can have a predictable result rather than an arbitrary drop of precision.

Only change was the source table was moved from Dictionary managed tablespaces to locally managed tablespaces. Details. PX 6.1., HP-UX, Oracle 8.1.7 We have a PX job "LoadETLDMSalePX" that writes to a table ETL_DM_SALE with the "Append" option. The other options are "Index M...

Copy the sql generated by the "Insert new or update Existing rows" option. Then change to "User defined sql". Paste the sql from above but make the changes to the update part to use the columns from the table instead of the host variables. This same mechanism can be used even if ...

What are the options you have selected for Source and Connection.

Have you selected "Partition Table"? If so PX needs to be able to read some of the sys tables to get partition information.