Multiple job has to write to a single file

[Maximus_Jack]

Hi
I have a requirement where multiple instance of a job has to write to a single file, i tried with both sequential file and dataset,

with sequential file- Somehow the records are missing when the job finishes..

with dataset - Sometimes I'm getting an error as "Error updating
ORCHESTRATE File Dataset descriptor for" this, sometimes i dont even get an error and the records wont be available in the file

is there any other way where multiple jobs write to a single file

Datastage version : 8.7 Fix pack 1

MJ

[chulett]

Simultaneously? No.

[Maximus_Jack]

thanks chullet..


strange... is there no way do it.. ?

[chulett]

It's not strange, it's the nature of sequential media which supports multiple readers but only a single writer. You'd need to sequence your jobs so they append their output.

[Maximus_Jack]

thanks...

but it should work for dataset isn't ?

[ray.wurlod]

No, because each segment file of a Data Set is a sequential file (same argument as before) and each parallel job has every segment file in the Data Set in use.

[Maximus_Jack]

thanks ray and chullet....