Comment #6 on issue 99 by mrd000: Memcached 1.4.2 server segmentation fault
http://code.google.com/p/memcached/issues/detail?id=99
I've hit the same problem now on 1.4.5. The patch above seemed to fix the
problem.
Server was crushing after ~ 15 minutes. We have a script which starts to
fill memcached actively.
Debug output was:
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0xb6baeb90 (LWP 29069)]
0x00365627 in event_add () from /usr/lib/libevent-1.4.so.2
#0 0x00365627 in event_add () from /usr/lib/libevent-1.4.so.2
#1 0x08049bb0 in update_event (c=0x80c31f0, new_flags=20) at
memcached.c:3313
#2 0x080520c6 in drive_machine (fd=35, which=2, arg=0x80c31f0) at
memcached.c:3394
#3 event_handler (fd=35, which=2, arg=0x80c31f0) at memcached.c:3732
#4 0x00365188 in event_base_loop () from /usr/lib/libevent-1.4.so.2
#5 0x080559f6 in worker_libevent (arg=0x8066368) at thread.c:245
#6 0x00b91832 in start_thread () from /lib/libpthread.so.0
#7 0x00b0fe0e in clone () from /lib/libc.so.6
I applied the patch and now it works for 1 hour.
