6

Got a head-scratcher here. Either I've found a Windows and/or VMware bug, or I've missed something stupendously simple. [Spoiler - it was stupendously simple.]

Our VMware environment is ESXI 5.5. We have a Windows 2012 R2 VM (named BOS-NETMON) that can no longer do SNMP queries of Cisco ASA devices (v2 or v3).

This machine runs SolarWinds Orion, and neither that nor standalone tools like Paessler SNMPTEST work. All other currently-monitored devices continue to respond to SNMP from this guest - we have Cisco IOS, Meraki, Exagrid, APC/Schneider - all fine. All other allowed protocols (SSH, HTTPS, ICMP) work when connecting to the ASAs from this guest.

The problem started when we switched the guest from the E1000e vNIC hardware to the VMXNet3. Prior to that, it worked fine for a couple of years.

  • Running the ASDM Logging window, it doesn't even see the attempted SNMP connection from BOS-NETMON. EDIT - This was because I didn't have the right logging settings on the ASA.
  • Running Wireshark on BOS-NETMON, it shows the SNMP queries going out, doesn't register responses from the ASAs.
    • I've tested SNMP from another VMXNet3 guest, it also fails to query SNMP against an ASA but works against non-ASA Cisco devices. EDIT - this doesn't appear to be true. Either I tested wrong, or it started working recently. Either way, another host with a VMXNet3 NIC can query SNMP against one of these ASAs.
  • I've tested SNMP from a guest with the E1000e vNIC, and it successfully queries SNMP against an ASA.
  • 4/5 of the target ASAs are not at the same site, so it shouldn't be an ARP issue - and if it were, the other non-SNMP protocols would be affected.

Further edit: I have ASDM debug info for a successful, and an unsuccessful, SNMP session. Next step will be a packet capture, I guess.

6   Mar 12 2018 16:30:26    302015  10.50.100.177   63809   10.10.99.10 161 Built inbound UDP connection 610885144 for Inside_Interface:10.50.100.177/63809 (10.50.100.177/63809) to identity:10.10.99.10/161 (10.10.99.10/161)

7   Mar 12 2018 16:30:26    710005  10.50.100.152   49588   10.10.99.10 161 UDP request discarded from 10.50.100.152/49588 to Inside_Interface:10.10.99.10/161
3
  • So it doesn't work with both local and remote ASA's?
    – joeqwerty
    Mar 12, 2018 at 20:02
  • While doing a port mirror on the ASA, do you see the incoming SNMP request? What does the ASA show as a response? Can you track this through the network and make sure something else didn't change that's now trying to drop these packets?
    – Andrew
    Mar 12, 2018 at 20:05
  • @joeqwerty - correct.
    – mfinni
    Mar 12, 2018 at 20:09

1 Answer 1

3

All other allowed protocols (SSH, HTTPS, ICMP) work when connecting to the ASAs from this guest.

Have you turned on snmp debugging on the ASA so you can see what it thinks?

debug snmp
logging buffered debug
logging asdm debug

Did the IP of the monitoring box change when you changed the vNIC? Are you using SNMP ACLs?

3
  • 1
    Aha. I am not the one who changed the vNIC, and the guy that did, fat-fingered the IP address when he put it back in.
    – mfinni
    Mar 12, 2018 at 20:45
  • 2
    TRUST NO ONE </xconfigfiles>
    – MikeyB
    Mar 12, 2018 at 20:53
  • 1
    SRSLY. SRSLY. So angry.
    – mfinni
    Mar 12, 2018 at 20:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.