SNMP Traps with no IP address

[robdennis]

I have two Veeam installations that I am trying to configure to notify our Nagios system. One is working ok, but for some reason the other is sending traps with a host IP address of 0.0.0.0. I can't find anything wrong in the setup, and don't know where the system would be looking to determine its IP address. Has anyone seen this problem before or have any idea as to where I should look to correct this problem.

[Vitaliy S.]

Can you try any other SNMP receiver to see if you have the same problem?

[robdennis]

I have three separate installations of B/R 6.5, two of them send SNMP traps correctly, but one of them does not supply the machine's IP address. It leaves the field blank with just zeros, so when I receive the trap, the system uses the WAN IP address of the senders network. The machine has a single static private IP address, just like the machines that are working properly. Below are examples of the two traps captured using Wireshark on the machines running Veeam.
The machine having the problem is running Windows Server 2008. I know that Veeam generates its own traps without using the Microsoft SNMP service, but I can't find where it would be getting (or not getting) the address from. Does anyone have any ideas what would be causing this?

Code: Select all

Simple Network Management Protocol
    version: version-1 (0)
    community: ********
    data: trap (4)
        trap
            enterprise: 1.3.6.1.4.1.31023.1.1.1 (iso.3.6.1.4.1.31023.1.1.1)
            agent-addr: 0.0.0.0 (0.0.0.0)
            generic-trap: enterpriseSpecific (6)
            specific-trap: 1
            time-stamp: 430012
            variable-bindings: 4 items

Code: Select all

Simple Network Management Protocol
    version: version-1 (0)
    community: ********
    data: trap (4)
        trap
            enterprise: 1.3.6.1.4.1.31023.1.1.1 (iso.3.6.1.4.1.31023.1.1.1)
            agent-addr: 192.168.0.23 (192.168.0.23)
            generic-trap: enterpriseSpecific (6)
            specific-trap: 1
            time-stamp: 6320593
            variable-bindings: 4 items

[Vitaliy S.]

It leaves the field blank with just zeros, so when I receive the trap, the system uses the WAN IP address of the senders network. The machine has a single static private IP address, just like the machines that are working properly.

What about other two Veeam B&R installations? Do you have a similar network setup in all three cases? Did you deploy Veeam B&R (the one that has problems) into a VM?

[robdennis]

The networks are all similar, but I am doing the captures on the machines running B/R so the network hasn't really come into play yet. They are all VMs. The ones that are working (now 3 separate installations) are on Win7 and WinXP machines. The one having issues is the only one running on a Win2k08 server.

[foggy]

Seems to be Microsoft issue (related to specific OS's, including w2k08): http://support.microsoft.com/kb/967181.

[robdennis]

Seems to be Microsoft issue (related to specific OS's, including w2k08): http://support.microsoft.com/kb/967181.

Foggy, THANK YOU!

That did fix it.
Since Veeam wasn't using the MS SNMP service (and the SNMP service was working correctly), I wasn't looking for a MS solution.
Applied the hotfix, rebooted, and the traps are now sent correctly.

Thanks Again.