On a xen dom0 (tested just now with 2.6.18-1.2869.fc6xen.x86_64), when snmpd is running, it barfs the following two lines into syslog every 30 seconds. Jan 8 12:47:00 barleycorn snmpd[2230]: netsnmp_assert index == tmp failed if-mib/data_access/interface.c:467 _access_interface_entry_save_name() Jan 8 12:47:00 barleycorn snmpd[2230]: netsnmp_assert __extension__ ({ size_t __s1_len, __s2_len; (__builtin_constant_p (rowreq_ctx->data.ifentry->name) && __builtin_constant_p (ifentry->name) && (__s1_len = strlen (rowreq_ctx->data.ifentry->name), __s2_len = strlen (ifentry->name), (!((size_t)(const void *)((rowreq_ctx->data.ifentry->name) + 1) - (size_t)(const void *)(rowreq_ctx->data.ifentry->name) == 1) || __s1_len >= 4) && (!((size_t)(const void *)((ifentry->name) + 1) - (size_t)(const void *)(ifentry->name) == 1) || __s2_len >= 4)) ? __builtin_strcmp (rowreq_ctx->data.ifentry->name, ifentry->name) : (__builtin_constant_p (rowreq_ctx->data.ifentry->name) && ((size_t)(const void *)((rowreq_ctx->data.ifentry->name) + 1) - (size_t)(const void *)(rowreq_ctx->data.ifentry->name) == 1) && (__s1_len = strlen (rowreq_ctx->data.ifentry->name), __s1_len < 4) ? (__builtin_constant_p (ifentry->name) && ((size_t)(const void *)((ifentry->name) + 1) - (size_t)(const void *)(ifentry->name) == 1) ? __builtin_strcmp (rowreq_c Wadda mess! It totally ruins the S/N ratio in /var/log/messages. And just FYI, here is a diff of the interface names/changes with and without xen kernel. --- ip-addr.2007-01-08 2007-01-08 12:13:33.000000000 -0800 +++ - 2007-01-08 12:44:00.105480000 -0800 @@ -1,10 +1,28 @@ 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 scope host lo -2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast qlen 1000 - link/ether 00:04:23:d0:53:cc brd ff:ff:ff:ff:ff:ff - inet 10.10.10.202/29 brd 10.10.10.207 scope global eth0 +2: peth0: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast qlen 1000 + link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff 3: eth1: <BROADCAST,MULTICAST> mtu 1500 qdisc noop qlen 1000 link/ether 00:04:23:d0:53:cd brd ff:ff:ff:ff:ff:ff 4: eth2: <BROADCAST,MULTICAST> mtu 1500 qdisc noop qlen 1000 link/ether 00:13:72:fe:3c:56 brd ff:ff:ff:ff:ff:ff +5: vif0.0: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 qdisc noqueue + link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff +6: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue + link/ether 00:04:23:d0:53:cc brd ff:ff:ff:ff:ff:ff + inet 10.10.10.202/29 brd 10.10.10.207 scope global eth0 +7: vif0.1: <BROADCAST,MULTICAST> mtu 1500 qdisc noop + link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff +8: veth1: <BROADCAST,MULTICAST> mtu 1500 qdisc noop + link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff +9: vif0.2: <BROADCAST,MULTICAST> mtu 1500 qdisc noop + link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff +10: veth2: <BROADCAST,MULTICAST> mtu 1500 qdisc noop + link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff +11: vif0.3: <BROADCAST,MULTICAST> mtu 1500 qdisc noop + link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff +12: veth3: <BROADCAST,MULTICAST> mtu 1500 qdisc noop + link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff +13: xenbr0: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 qdisc noqueue + link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff
Can you please try to reproduce this issue with the latest rawhide net-snmp? I can't reproduce the above assertion on my Xen test box.
Not reproducible with 5.4-5.fc7.x86_64. This bug appears in fc6 i.e. net-snmp 5.3.1.
Awwww foo! I roll back to fc6 package and now bug is gone. However... # chkconfig snmpd on # shutdown -r now It's baaaack. Even further, upgrade to 5.4 (this time my mirror had the rawhide package you built last night), reboot and the problem persists. So it is reproducible with both 5.3 and 5.4. Just only when started via boot. Is there a race here? Or something screwey with my environment?
Altough my messages in /var/log/messages were quite different, the fact is, that the snmpd logs wierd things. This difference may be configuration related, but it is bug anyway. The bug is reporducible.
Fixes checked in upstream for 5.3.x: -fix (reported elsewhere) xen+x86_64 crash: http://net-snmp.cvs.sourceforge.net/net-snmp/net-snmp/snmplib/snmp_logging.c?r1=5.34.2.5&r2=5.34.2.1&pathrev=V5-3-patches -fix for overly-verbose log message: http://net-snmp.cvs.sourceforge.net/net-snmp/net-snmp/agent/mibgroup/if-mib/ifTable/ifTable_data_access.c?r1=1.18.2.3&r2=1.18.2.4
Both pathes are in the latest rawhide version, I believe this issue is fixed in net-snmp-5.4-13
It still appears in F8 net-snmp-5.4.1-5.fc8 Dec 4 21:15:12 *** snmpd[32738]: netsnmp_assert index == tmp failed if-mib/data_access/interface.c:469 _access_interface_entry_save_name() @owner: please change version to 8 and reopen this bug. I got this message after a while, kill -HUP doesn't help: Dec 4 21:21:32 * snmpd[32738]: Reconfiguring daemon Dec 4 21:21:32 * snmpd[32738]: NET-SNMP version 5.4.1 restarted Dec 4 21:21:42 * snmpd[32738]: netsnmp_assert index == tmp failed if-mib/data_access/interface.c:469 _access_interface_entry_save_name() Dec 4 21:22:15 * snmpd[32738]:last message repeated 2 times Dec 4 21:22:15 * snmpd[32738]: Received TERM or STOP signal... shutting down... Dec 4 21:22:16 * snmpd[14238]: netsnmp_assert !"registration != duplicate" failed agent_registry.c:535 netsnmp_subtree_load() Dec 4 21:22:17 * snmpd[14238]:last message repeated 2 times Dec 4 21:22:17 * snmpd[14238]: error finding row index in _ifXTable_container_row_restore Dec 4 21:22:17 * snmpd[14238]: error finding row index in _ifXTable_container_row_restore Dec 4 21:22:17 * snmpd[14238]: NET-SNMP version 5.4.1 Finally I found the reason: Nov 26 12:39:03 * pppd[15672]: primary DNS address *** Nov 26 12:39:03 * pppd[15672]: secondary DNS address *** Nov 26 12:39:04 * snmpd[11638]: netsnmp_assert index == tmp failed if-mib/data_access/interface.c:469 _access_interface_entry_save_name() Nov 26 12:39:19 * snmpd[11638]: netsnmp_assert index == tmp failed if-mib/data_access/interface.c:469 _access_interface_entry_save_name() Problem is caused by changing of ppp interface after hangup and re-dialin. This problem leads also to an mrtg problem, where mrtg can no longer retrieve any interface data (I've tried already \ppp@...) I believe the internal reason is that the ppp interface gets a new index number, this disturbs net-snmp and as secondary reason, mrtg get no data anymore. Workaround: Append to /etc/ppp/ip-up.local: /sbin/service snmpd restart >/dev/null