Originally posted by: shargus
What good is setting limits if you cannot tell from a system level that they are being violated? Well... the limits are enforced by the kernel. If a process requests a resource that exceeds those limits, the request is denied and an error code is returned. In theory, the process can try again with a smaller request.
So, the real answer would be - a process cannot violate the limits set by the kernel.
Is there a log for when a process
attempts to violate the limits? That would probably be up to the application that makes the attempt.
Beyond that - maybe something with the auditing system... ?