It's possible the kernel is reaching some sort of sysctl limit.
My suggestion would be to dump the full process list at the time of occurrence, examine the processes one by one to see if you can find any information and examine the sysctl values on your machine.
It's possible the kernel is reaching some sort of sysctl limit.
My suggestion would be to dump the full process list at the time of occurrence, examine the processes one by one to see if you can find any information and examine the sysctl values on your machine.