We have several cases where a scheduled job like aggregation or an audit process fails when run at scheduled times, but succeed later on when run manually.
The most common cause for these is some environmental factor like exhausted memory or some other job impeding this one. These are somewhat hard to catch sometimes with an app or agg debug turned on at that time. But this is cumbersome, since many of these run at off-hours when the systems may not be manned.
Requesting a feature where we add a "debug flag" to the scheduler dialog box that the customer can turn on/off.
When turned on, just prior to when the job is triggered, we would kick off app or agg debug as necessary, along with a "snapshot" of the running environment like memory, other jobs running, mysql running processes etc. This will give Support valuable information in order to find the cause of issues much quicker and in a more reliable manner.
Once we have have this information, the customer can then turn debug off.
Do not place IBM confidential, company confidential, or personal information into any field.