Every 20 minutes, the controller goes down for one of HP-UX 11.23 ia64 robot.
Nothing is recorded in the controller.log file, but we see the following recorded in the nimbus.log on this robot:
Aug 1 09:42:11:656 Nimbus: 'controller' stopped (pid=11014) SIG = 9 Aug 1 09:42:11:657 Nimbus: Too many retries used starting 'controller', next start in 1200 seconds
SIG = 9 corresponds to a core dump on HP-UX UNIX systems.
The controller fails to start, then waits 20 minutes and tries again, but continues to fail with core dumps.
Cause
Execute permission was removed from the shared libraries installed in the <installPath>/nimsoft/robot/lib directory:
-rw-r--r-- 1 root root 68368 May 4 13:48 libmodule_loader.sl -rw-r--r-- 1 root root 756000 May 4 13:48 libnim_shared.sl
A shared library with mode 644 on an HP-UX system will cause core dumps.
Environment
UIM 8.51 robot: 7.80 OS: HP-UX 11.23 ia64
Resolution
Change permissions of the shared libraries used by the controller probe and add execute permission: