Menu
▾
▴
[ net-snmp-Bugs-1644670 ] 5.4 snmptrapd crashed after forwarding SNMPv3 traps
|
From: SourceForge.net <no...@so...> - 2012-10-24 20:04:15
|
Bugs item #1644670, was opened at 2007-01-25 10:36 Message generated for change (Comment added) made by auralox You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=112694&aid=1644670&group_id=12694 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: snmptrapd Group: None Status: Open Resolution: None Priority: 5 Private: No Submitted By: Chau Trinh (chautrinh) Assigned to: Nobody/Anonymous (nobody) Summary: 5.4 snmptrapd crashed after forwarding SNMPv3 traps Initial Comment: I have 3 i686 machines, running Linux 2.6.16.16 with net-snmp-5.4, A, B, and C. B is configured to forward traps to C. snmptrapd is running interactively on B and C. At A, a v3 trap was sent to B, using snmptrap command. The trap was received successfully at C but snmptrapd crashed at B. Here is the backtrace: NET-SNMP version 5.4 2007-01-24 15:37:20 <UNKNOWN> [UDP: [192.168.22.2]:32968]: DISMAN-EVENT-MIB::sysUpTimeInstance = Timeticks: (42) 0:00:00.42 SNMPv2-MIB::snmpTrapOID.0 = OID: SNMPv2-SMI::enterprises.1569.223.1.1.6.0.4 SNMPv2-SMI::enterprises.1569.223.1.1.5.2.0 = INTEGER: 2 Program received signal SIGSEGV, Segmentation fault. [Switching to Thread -1215118432 (LWP 21163)] 0xb79d6e84 in mallopt () from /lib/tls/libc.so.6 (gdb) bt #0 0xb79d6e84 in mallopt () from /lib/tls/libc.so.6 #1 0xb79d5dcb in free () from /lib/tls/libc.so.6 #2 0xb7d5c66c in usm_free_usmStateReference (old=0x80759d8) at snmpusm.c:189 #3 0xb7d2c2b9 in _sess_process_packet (sessp=0x80aaba8, sp=0x8065068, isp=0x8077a78, transport=0x809bd98, opaque=0x80a7bd8, olength=20, packetptr=0x80b1778 "0\201¦\002\001\0030\021\002\004<*±@\002\003", length=169) at snmp_api.c:5376 #4 0xb7d2d4fb in _sess_read (sessp=0x80aaba8, fdset=0x8077a78) at snmp_api.c:5779 #5 0xb7d2dcc9 in snmp_sess_read (sessp=0x80aaba8, fdset=0x1011f170) at snmp_api.c:5798 #6 0xb7d2c94c in snmp_read (fdset=0xbfc4e4b0) at snmp_api.c:5412 #7 0x0804cee9 in main (argc=5, argv=0xbfc4e654) at snmptrapd.c:1418 ---------------------------------------------------------------------- Comment By: Auralox (auralox) Date: 2012-10-24 13:04 Message: I am having the same problem with a SunOS 5.10 machine running net-snmp 5.3 forwarding to a RHEL 6 machine running net-snmp v5.5. The machine can properly receive the traps and does in fact forwards the 1st one on, but the snmptrapd crashes ungracefully with a SIGSEV immediately following the packet being sent. This is a hugely problematic bug for anyone trying to implement secure (i.e. v3) snmptrap forwarding - in my case from a console network on a non-routable subnet to our monitoring system on the WAN. I've also tried to adjust the context name to be greater then 32 as recommended by zuuooo but with no luck on the 5.3.1 version available under solaris 10. Anyone capable of taking this on? It is marked as High-Priority but has not been assigned to anyone. It would be really awesome to have this resolved - I am sure other folks need this functionality as well. Seems that this bug has been around for over 5 years... so don't know what the problem is. ---------------------------------------------------------------------- Comment By: zuuooo (zuuooo) Date: 2012-06-19 12:40 Message: I have the same situation, this happens only when contextName length is smaller than 32. Temporary solution is to use contextName with length grater or equal 32. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=112694&aid=1644670&group_id=12694 |
