Better way to clear stuck reports?

Afternoon all,

We semi-regularly encounter stuck reports of all types. Attempting to cancel them through the System Monitor has no effect and we need to open a ticket with Cloud Support. Depending on the rep I get, this process either goes very quickly or can take hours. During this period of time and depending the report, users can be significantly impacted.

Has anyone found of a way to take care of this issue themselves? Why does it happen so often? It’d be great if Epicor wrapped up whatever steps the Cloud team take on our server and made it accessible to users.

Not really specific to reports but whenever something is stuck, at least in dev it happens when we restart the application pool while a task is running, so it stays there but no longer doing anything. When you click cancel in system monitor it creates a record in Ice.SysTaskKill, which processes/reports periodically check to abort, but since nothing is running nothing happens.

What we do is restart the task agent, this moves the stuck record to the history tab with status CANCELLED.

What impact are you seeing from stuck reports?

We are cloud, so (I believe) we have no ability to restart the application pool or task agent. These are random reports that sporadically stick in System Monitor, blocking the user from running that same report again. This kills whatever workflow they were performing.

I would open a separate ticket as to why the report keeps getting stuck. Unless there is some weird exception happening it should not, but without looking into the server and logs it is hard to tell.

Is there any pattern? Same report, parameters, user, time of day, etc?

As for when it happens, it might be faster to just cancel them all through System Monitor and then create a ticket specifically to restart the Task Agent, maybe being more direct makes it easier for the rep and therefore faster, maybe some reps take a bit more time if they don’t know what to do when “things are stuck”.

There hasn’t been any obvious pattern. Financial reports, purchasing reports, the job traveler. This has happened mornings, it’s happened afternoons. It’s happened to several reports at the same time, or just one.

Needing to restart the app pool is an extremely common thing, whether it’s your own mistake or an Epicor bug that caused something to get stuck. There’s an open suggestion on the Epicor Ideas board to make restarting the app pool accessible to users.

Meanwhile, we’re dropping cloud.