Re:Killing processes that don't want to die
by taejo on Tuesday September 25, 2018 @07:41AM
(#18065416)
Our product occasionally does this at the moment. We've set up asserts in our application to launch a "phone-home" process (on the theory that launching a new process is more likely to succeed than trying to phone home from the borked process). However, the phone-home program uses the same assert library, so the rare assertion failure in the phone-home will launch a new phone-home.
[ Parent
| on HN ]
There are no comments on this comment yet.