Anyone ever seen a process fork but then the child hangs so never gets to the exec line? fork returns a valid pid and the process does get created since our app has a timeout and it catches the hung subproc at shutdown.
There is nothing relevant in /var/adm/messages.
There is a cout in the forked process before the exec which did not produce any output. Maybe this is not a good idea to have a cout in the forked process since it uses the file handle from the main thread?
Our app has been running for months and we haven't seen this before. Can't reproduce it either. There is a known solaris issue which we have fixed where processes exec and never start. We fixed this by using the thread stuff from /usr/lib/lwp. The app forks hundreds of subprocs during a batch, about 16 at a time. The machine is pretty heavyweight with 8Gb of memory and 8 processors. Running Solaris 2.8.
Maybe I'll use the old excuse that it was a stray cosmic particle....
========================================
hmm - should I concentrate on C# or Java next?....
========================================
There is nothing relevant in /var/adm/messages.
There is a cout in the forked process before the exec which did not produce any output. Maybe this is not a good idea to have a cout in the forked process since it uses the file handle from the main thread?
Our app has been running for months and we haven't seen this before. Can't reproduce it either. There is a known solaris issue which we have fixed where processes exec and never start. We fixed this by using the thread stuff from /usr/lib/lwp. The app forks hundreds of subprocs during a batch, about 16 at a time. The machine is pretty heavyweight with 8Gb of memory and 8 processors. Running Solaris 2.8.
Maybe I'll use the old excuse that it was a stray cosmic particle....
========================================
hmm - should I concentrate on C# or Java next?....
========================================
