I would like to see an enhancement either to JMX notifications, or through PSM 3.0 notifications to notify us of increased failure rate for publishing jobs.
This is the problem we face.
An error occurs with a cad worker or office worker, or delegate. The error results in all WVS publishing jobs assigned to that worker failing. In essence the worker becomes an automated job squashing machine. (In my mind I picture the Monty Python Foot squashing them into wine or something.) This makes users unhappy, and if you work in a diverse publishing environment with some minimal level of failures occurring regularly it might not be noticed by the administrator(s) for some time.
I think this would allow us to respond to these issues faster.
I would like to see a monitor similar to the response time monitors in PSM 3.0 that show when the average request time crosses some threshold. Only I would like to see it track failure rate during some time frame, either by queue, queueset, or by worker. Then when the failure rate crossed some threshold administrators would get an e-mail notification so they could investigate it before a user calls to complain.
I have seen problems of this type with Creo workers, Solidworks workers, and Inventor Workers, and most recently with office workers using Adobe LiveCycle ES4.
If anyone knows of a way to configure a notification for this I would be interested in that too.