I haven't tried this method yet, I'll have to test it out. What I do is
just set a Static 0.0.0.0/0 route with the gateway on each Backhaul with
a different routing mark.
firewall/mangle
mark Connection, chain Input, UDP port 161, set In Interface of Backhaul
Mark Routing, Chain Output, Check Connection mark from above.
2 rules for Each backhaul link Since it's only looking at Input 161, no
extra CPU overhead.
On 10/19/2017 3:13 PM, Mike Hammett wrote:
SRC NAT the SNMP port to the loopback IP.
-----
Mike Hammett
Intelligent Computing Solutions <http://www.ics-il.com/>
<https://www.facebook.com/ICSIL><https://plus.google.com/+IntelligentComputingSolutionsDeKalb><https://www.linkedin.com/company/intelligent-computing-solutions><https://twitter.com/ICSIL>
Midwest Internet Exchange <http://www.midwest-ix.com/>
<https://www.facebook.com/mdwestix><https://www.linkedin.com/company/midwest-internet-exchange><https://twitter.com/mdwestix>
The Brothers WISP <http://www.thebrotherswisp.com/>
<https://www.facebook.com/thebrotherswisp>
<https://www.youtube.com/channel/UCXSdfxQv7SpoRQYNyLwntZg>
------------------------------------------------------------------------
*From: *"George Skorup" <[email protected]>
*To: *"Animal Farm" <[email protected]>
*Sent: *Thursday, October 19, 2017 3:12:12 PM
*Subject: *[AFMUG] Stupid MikroTik SNMP
I know Nate has mentioned this before. Wondering if there's a solution.
This has been driving me nuts for years.
A large/complicated OSPF design may have some asymmetric paths between A
and Z. But the problem comes down to asymmetry at the router you're
trying to poll.
For example, SNMP polling a router to its loopback IP, requests come in
on say ether1, but the replies go out ether2 = SNMP timeout. In and out
same interface works fine.
Everything else to the router works fine, like WWW, telnet, winbox, etc.
but obviously those are TCP, so this has me wondering if it's a UDP
thing or just SNMP...?
