We are a small ISP just getting into Canopy, so please forgive me if this is not quite right!
We currently have a few units (P9 … 5.7 SM’s, 2.4 10MB BH, 5.7 AP’s) and they are set up and working fine with a bridge timeout of 6 hrs (not sure if this could be longer…) and a public 169.254 addressing. We also have protocol filtering in place. (SMB, SNMP, Bootp Server, IPv4 Multicast)
We go through a Mikrotik router and our IP address range (203.xxx) is across two ethernet cards (for static IP addressing and the other is PPPoE). Problem we have is that when the interface cards on the router are set to ARP enabled - we can see the Canopy units. We set the router to proxy-arp to allow our own servers on the same IP address range then we loose the ability to connect to the Canopy units - yet SNMP still works ok and Cacti is still graphing the units fine.
This has me confused - I’m sure it’s one of those D’OH questions - any help would be greatly appreciated!
Don’t you need static ARP entries on your router at this point for your Canopy equipment?
The SNMP equipment uses layer 3 so it’s probably building it’s own ARP tables dynamically using broadcast on the same subnet.
Ok, I have added the static ARP entries into the router. Now the network seems more unstable?
We can sometimes ping the units, othertimes they are unavailable. Also connectivity seems to drop in and out. Before Internet worked fine … now this seems to be affected and Cacti graphs are showing an increase in traffic level after the entries went into the router (where there was not traffic before) as well as units dropping off then on
Curious to know what other users do when setting up proxy-arp (if they do at all) … for their Canopy units.