Red Hat Bugzilla – Bug 202393
Intermittent named crash with pthread_mutex_destroy
Last modified: 2013-04-30 19:34:00 EDT
Description of problem:
Very occasionally (particularly under load) named will crash with one ofthe
following (or similar)
RUNTIME_CHECK(((pthread_mutex_destroy(((&adb->mplock))) == 0
) ? 0 : 34) == 0) failed
].lock))) == 0) ? 0 : 34) == 0) failed
I don't believe this is related to memory constraints, since the busiest server
never exceeds a resident footprint of 250MB, whereas these servers have 2GB RAM
and are exclusively for DNS.
Version-Release number of selected component (if applicable):
Kernel 2.6.9-34.0.2.ELsmp x86_64
Happened about four times in eight months (so not very :< )
Steps to Reproduce:
1. Run bind on an SMP x86_64 box.
This is essentially the same issue as bug 173961, which was closed as it claims
a fix has been implemeneted in back-ports from Bind 9.2.6 to 9.2.4, but we are
*** Bug 202394 has been marked as a duplicate of this bug. ***
bind-9.3.2-2_EL4 isn't supported by Red Hat, it's only a test package...
This is fixed in RHEL-4.4 so if you want use bind 9.3 series or newer please use
packages from Fedora. bind 9.3.2 isn't supported in RHEL-4 => closing
*** This bug has been marked as a duplicate of 173961 ***