SNMP Discovery fails with 2.2.4

[lostinspace2011]

I recently upgraded from 2.0.12 to 2.2.4, however since upgrading some of the hosts no longer work. It seems that SNMP discovery has changes between these version.

What seems to be happening is that the discovery kills the SNMP service on the target host/device. I understand this is not a ZABBIX issue as such, and I have already raised a support case with the device vendor.

I am hopeing however that somebody can confirm there were some changes to the SNMP discovery and to help me get the commands used by zabbix to make these discoveries. This would help me supply the device vendor with details on how to recreate the problem, without installing zabbix.

Both commands run successfully
snmpwalk -v2c -c public 192.168.0.1
snmpwalk -v2c -c public 192.168.0.1 -On

[lostinspace2011]

Error from router

The error reported on the router is:

Jul 15 13:27:02 daemon warn kernel: snmpd/8413: potentially unexpected fatal signal 11.
Jul 15 13:27:02 daemon warn kernel: Cpu 1
Jul 15 13:27:02 daemon warn kernel: $ 0 : 00000000
Jul 15 13:27:02 daemon info kernel: 00000001 09044301 00000007
Jul 15 13:27:02 daemon warn kernel: $ 4 : 7f9018a8
Jul 15 13:27:02 daemon info kernel: 00000000 00000000
Jul 15 13:27:02 daemon info kernel: 7f9018a8
Jul 15 13:27:02 daemon warn kernel: $ 8 : 00000000
Jul 15 13:27:02 daemon info kernel: 00000000
Jul 15 13:27:02 daemon info kernel: 00000006 00000001
Jul 15 13:27:02 daemon warn kernel: $12 : 00000002
Jul 15 13:27:02 daemon info kernel: 00000001 00000002
Jul 15 13:27:02 daemon info kernel: 00000002
Jul 15 13:27:02 daemon warn kernel: $16 : 7f9027f4 00000003
Jul 15 13:27:02 daemon info kernel: 004028a0 2af42000
Jul 15 13:27:02 daemon warn kernel: $20 : 00402320 2af20230 7f902738
Jul 15 13:27:02 daemon info kernel: 00407e70
Jul 15 13:27:02 daemon warn kernel: $24 : 00000000 2af20370
Jul 15 13:27:02 daemon info kernel:
Jul 15 13:27:02 daemon warn kernel: $28 : 00435050
Jul 15 13:27:02 daemon info kernel: 7f9018b8
Jul 15 13:27:02 daemon info kernel: 7f9018b8
Jul 15 13:27:02 daemon info kernel: 00408e9c
Jul 15 13:27:02 daemon warn kernel: Hi : 08888768
Jul 15 13:27:02 daemon warn kernel: Lo : cd29f007
Jul 15 13:27:02 daemon warn kernel: epc : 00408f1c 0x408f1c
Jul 15 13:27:02 daemon warn kernel: Tainted: P
Jul 15 13:27:02 daemon warn kernel: ra : 00408e9c 0x408e9c
Jul 15 13:27:02 daemon warn kernel: Status: 00008d13
Jul 15 13:27:02 daemon info kernel: USER EXL
Jul 15 13:27:02 daemon info kernel: IE
Jul 15 13:27:02 daemon warn kernel: Cause : 00000010
Jul 15 13:27:02 daemon warn kernel: BadVA : 09044309
Jul 15 13:27:02 daemon warn kernel: PrId : 0002a070 (Broadcom4350)

[lostinspace2011]

Almost working

Switching from SNMP2 to SNMP1 improved things a little. There were still issues with some of the items. Looks like the router doesn't handle the parsing of HEX encoded IP addresses well. I ended up removing those items from the template and now it is all mostly working.

Thanks to the guys on IRC, especially LizardKing for helping me investigate this issue.

[lostinspace2011]

Unable to reproduce using the Zabbix 2.2.2 appliance

To be able to supply the router vendor with the steps to reproduce the problem I figured it might be good to use the zabbix appliance and describe the configuration steps.

Well, I documented the steps, but can't seem to reproduce the issue. The zabbix server on which I am having the issue is running Gentoo and Zabbix 2.2.4 while the appliance is running Suse and Zabbix 2.2.2.

This leads me to believe the problem which caused snmpd on the router to crash could be on the zabbix server side.

Are there any commands I can run on the zabbix server to reproduce the discovery behaviour. ?