This is already happening. The user selects a too complex pivot table, it takes a long time, memory is full, the service is cleared and all users suffer (about 200 people just go for a walk).
The mechanism specified in the screenshot (object calculation time limit), as I understand it, works on a thread (or core) I do not remember, but I think that this is idiocy.
This mechanism is useless, it is generally inappropriate.
When will you make a normal mechanism for limiting the calculation of an object by time, regardless of the time per core or thread?
Also, please consider the option with several services multiple of the number of reports in memory. If one report takes up too many resources, then the whole service and all reports break, if you use several services, then you could use the mechanism of dropping the service depending on the memory limitations on one service.
Then all users would cease to suffer.
