DSXchange

I am bringing in 15million+ rows into a join using an OraEnt stage. Job aborts with error messages such as "Unspecified error", "pipe write faile","fileReadAck()" etc. When I put the data of this stage into a dataset and use the dataset to do the join, the job works. An...

How are the commands dsrecords and orchadmin run in a Windows environment? Do I log in to command prompt and run them?

I have thoroughly searched this forum and other internet sites such as ittoolbox but couldn't find a whole lot on the specifics. Here are a few points I found: 1)Windows OS requires far more server reboots than Unix 2)Unix allows much larger file size handling 3)If source data is in Unix, DS on Unix...

I have 3 issues but all of them seem related so I am posting them together. 1) Performance statistics not showing for some jobs (it is showing for some others): Job design: 2 Ora Enterprise stages as sources. OraEnt1 goes to a modify and on to Lkp1. OraEnt2 goes to Lkp1 directly. Lkp1 goes to Transf...

That's exactly what I am saying. If my commit level is 5000 then I should be able to insert 5000 (which has a row with PK of 123) and I should be able to update 5000 (which also has a row with updates on PK of 123). If I am understanding Ray's post correctly, this value cannot be greater than 1.

Ray, so are you saying that Datastage sends a row through the insert link and then one through the update link and then nsert link? What if rows/transaction is 5000? I would think all 5000 rows would be sent in the sequence insert(5000)/update(5000)/insert(5000).

I am inserting data and updating data using output from the same transformer stage using 2 different OraOCI9 stages. In the Link Ordering, the insert is first and update is second. If I have a record coming in which needs to be inserted first and updated second, shouldn't the insert happen first and...

I can trap the value after the execution of the ExecCommand stage in the Triggers tab or in the next stage. However, by that time, the warning has been logged which I want to prevent.

Where would be the best way to trap the return value from the job so that I can prevent the warning?

Thanks Craig for your response. -jobstatus: Waits for the job to complete and returns an exit status -wait: Waits for the job to complete Can you tell me why a successful exit status of 1 be considered a warning in the log? I am not using a JobActivity stage because the job name is a parameter, i.e....

I am executing a job as: dsjob -run -jobstatus -param Prm1=Value1 ProJName JobName I am putting this in an ExecCommand stage in a job sequence. The job executes fine without any warnings or errors of any kind. However, the ExecCommand stage returns with a message in the Director: dsjob did not finis...

I saw your routines. It doesn't have one that works with a second and a timestamp. I guess I'll have to manipulate an existing one.

What is the best way to go about doing it?

By performance, I mean I want the entire sequence to finish as fast possible.

I have 30+ jobs and none of them are dependent on another. Some of them take a long time to finish while others finish in seconds. What would be the ideal way to place them in a job sequence so that performance for the overall sequence is best?

I put in double quotes around the -ape but the error is the same.