Hello, pardon my lack of experience with RouterOS (total noob here, this is my first post).
I have been running RouterOS on a server with 2 GigE ports - one used for management, the other for my hotspot services. At times, I run 6 different WLANs with or without hotspot services (all of them are different) which requires me to trunk a bunch of VLANs through it. I am very impressed with how easy this was to implement and how absolutely flawless this has been operating.
But, given that I have anywhere from 2000 to 5000 users on this network at any given time, I am afraid of a hardware failure of this server and I need to implement a redundancy solution ASAP. I built a lab where I have two of these servers handling all of the WLAN traffic from a Ruckus Zone Director 5K.
I have the RouterOS configured as follows:
6 VLANs with /16 networks Primary 10.1.0.2 Secondary 10.1.0.3 ip addressing
Each networks has a VRRP interface with 10.1.0.1/32
clients auth through radius
I have built the hotspot services on the VRRP interfaces on both sides.
This works great until I reboot the machines, then the VRRP instance seems to go berserk. Both servers are running in RM mode and it seems neither of them will speak to eachother. I thought this may be due to the hotspot service, not authenticating the other server, so I added them to the ACL, the walled garden, and poked half a million holes through the firewall for the VRRP protocol as well as port 112, but I cannot seem to find out where I am going wrong here.
Is there anyone else out there with this type of implementation that has any clues as to what I might be doing wrong?
Thanks,
//P