I do not believe the SAS note is the cause of the issue we experienced, but thanks for the hint. I'm intrigued by your statement "BTW, this is one of the reasons I would not use EG-based scheduling for business-critical work - it is inherently less reliable that server-based scheduling.". As I'm fairly new to this world, I'd love to hear any recommendations. To give you a better understanding of our environment, these jobs are running on a dedicated server and are scheduled using Control-M. There are 100's of daily jobs scheduled on this server. Many are straight-up SAS programs that are executed, some are ITRM jobs and others are processes within EGP launched using a VB Script. (pardon me if I've butchered any nomenclature) Generally these jobs run fairly smooth, with the occasional external-caused failure, however in this particular case, these EG jobs using the EGP VBScript simply stopped executing. Unfortunately, as the VBScript didn't pass back an error Control-M expected it erroneously flagged the job as completed, but that's a whole different can-of-worms. After many hours of troubleshooting these jobs (source data validation, connectivity, recent changes, etc.) we determined we could manually run the jobs from EG on our desktops and they executed successfully. Digging into the server and executing them as the scheduler account, along with altering the error-trapping code in the VBScript we finally were enlightened to the fact the Connection Profile didn't exist. We were able to simply create a new connection profile and the world was right again. The oddest part is, other jobs had run merely minutes prior successfully using the same VBScript/EG/Connection Profile, then suddenly, poof, profile gone. Thanks for any insight or recommendations. Jim
... View more