I have two different applications on two different hosts under FreeBSD
9.0: Munin and Cacti/spine. These applications have been running for
several months on the first and several years on the second.
Recently, after a patch cycle, both systems are running out of process
slots. From a "ps" perspective, zombies are not being reapped. For
example:
80525 ?? S 0:00.15 /usr/local/bin/spine 32 49
80527 ?? Z 0:00.10 <defunct>
80530 ?? Z 0:00.10 <defunct>
80532 ?? Z 0:00.09 <defunct>
80534 ?? Z 0:00.09 <defunct>
80536 ?? Z 0:00.09 <defunct>
80538 ?? Z 0:00.13 <defunct>
80539 ?? Z 0:00.14 <defunct>
I GDB the spine application and find:
Starting
program: /disk-2/ports/net-mgmt/cacti-spine/work/cacti-spine-0.8.8/spine
32 49
[New LWP 118050]
[New Thread 802c07400 (LWP 118050/spine)]
SPINE: Using spine config file [/usr/local/etc/spine.conf]
SPINE: Version 0.8.8 starting
[New Thread 802c09800 (LWP 121307/spine)]
[New Thread 802c09c00 (LWP 121309/spine)]
[New Thread 802c0a000 (LWP 121311/spine)]
[New Thread 802c0a400 (LWP 121313/spine)]
[New Thread 802c0a800 (LWP 121315/spine)]
[New Thread 802c0ac00 (LWP 121317/spine)]
[New Thread 802c0b000 (LWP 121319/spine)]
[New Thread 802c0b400 (LWP 121321/spine)]
[New Thread 802c0b800 (LWP 121323/spine)]
[New Thread 802c0bc00 (LWP 121325/spine)]
[New Thread 802c0c000 (LWP 121327/spine)]
[Thread 802c09800 (LWP 121307/spine) exited]
[New Thread 802c09800 (LWP 121307/spine)]
[New Thread 802c0c400 (LWP 118883/spine)]
[Thread 802c09c00 (LWP 121309/spine) exited]
[New Thread 802c0c800 (LWP 118891/spine)]
[Thread 802c0bc00 (LWP 121325/spine) exited]
[New Thread 802c0cc00 (LWP 118909/spine)]
[Thread 802c0ac00 (LWP 121317/spine) exited]
[New Thread 802c0d000 (LWP 121331/spine)]
[Thread 802c0c000 (LWP 121327/spine) exited]
[New Thread 802c0d400 (LWP 121333/spine)]
[Thread 802c0c400 (LWP 118883/spine) exited]
[New Thread 802c09800 (LWP 120582/spine)]
[Thread 802c0b000 (LWP 121319/spine) exited]
[New Thread 802c09c00 (LWP 120566/spine)]
[Thread 802c0b800 (LWP 121323/spine) exited]
[Thread 802c09800 (LWP 120582/spine) exited]
[Thread 802c0b400 (LWP 121321/spine) exited]
[Thread 802c0d400 (LWP 121333/spine) exited]
[Thread 802c0a000 (LWP 121311/spine) exited]
[Thread 802c09c00 (LWP 120566/spine) exited]
[Thread 802c0d000 (LWP 121331/spine) exited]
[Thread 802c0cc00 (LWP 118909/spine) exited]
[Thread 802c0c800 (LWP 118891/spine) exited]
[Thread 802c0a800 (LWP 121315/spine) exited]
[Thread 802c0a400 (LWP 121313/spine) exited]
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 802c07400 (LWP 118050/spine)]
0x00000008020f88b6 in _pthread_mutex_init_calloc_cb ()
from /lib/libc.so.7
(gdb) up
#1 0x00000008020fc07b in _malloc_postfork () from /lib/libc.so.7
(gdb)
#2 0x00000008020fc1be in _malloc_postfork () from /lib/libc.so.7
(gdb)
#3 0x00000008020fd048 in _malloc_prefork () from /lib/libc.so.7
(gdb)
#4 0x0000000802102059 in free () from /lib/libc.so.7
(gdb)
#5 0x0000000800ac0200 in netsnmp_ds_shutdown () at default_store.c:555
555 free(drsp->ftype);
(gdb) l
550
551 if (drsp->ftype && drsp->token) {
552 unregister_config_handler(drsp->ftype, drsp->token);
553 }
554 if (drsp->ftype != NULL) {
555 free(drsp->ftype);
556 }
557 if (drsp->token != NULL) {
558 free(drsp->token);
559 }
(gdb)
I don't see anything obvious in the net-snmp archives. Being new to this
list I would appreciate if someone would clue me in as to what changed
and what to look for.
------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Net-snmp-users mailing list
Net-snmp-users@lists.sourceforge.net
Please see the following page to unsubscribe or change other options:
https://lists.sourceforge.net/lists/listinfo/net-snmp-users
