Red Hat Bugzilla – Bug 584574
groupd blocks on startup when spemaphore limit exceeded.
Last modified: 2010-05-31 09:23:28 EDT
Description of problem:
One cluster startup if there are is no more semaphore space available groupd will block and not throw an error message.
strace of groupd shows:
semget(0x11b8c82e, 3, IPC_CREAT|IPC_EXCL|0600) = -1 ENOSPC (No space left on device)
geteuid() = 0
ipcs -a hows semaphore count is equal to kernel.sem.
Version-Release number of selected component (if applicable):
- every time
Steps to Reproduce:
- Use up all semaphores
- Attempt to start cluster suite
cman startup hangs
Cluster startup should error out with a useful message
Have a look at resolving this problem.
Also changing to 5.4 from 5.4.z since this is not a z stream regression.
This bug is duplicate of https://bugzilla.redhat.com/show_bug.cgi?id=561546 so closing this bug as duplicate. You can find solution based on POSIX semaphores in mentioned bug.
*** This bug has been marked as a duplicate of bug 561546 ***