Is it ok to make the US-16XG as the core switch? and distribution switch (TP-Link)? Compatibility?
What if the end device just have a gigabit network card? Or it’s just needed for future proof?
Only if you plan to route using the L3 switch and not on pfsense. However, my knowledge on l3 protocols on switches is limited. Perhaps someone can help on this. Generally if you use VLAN you would want some kind of control over traffic flow, hence where firewall rules come in.
But I on my core switch I only have 4Ports… If I use LACP to bundle 2links as 1 port that would still be enough for only connecting 2 Switches… In reality Switch 5 and 6 is really far to IT Switch that’s why we can’t add redundancy link…
This is the end devices, we are looking at the eventual traffic flow between switches up and downstream.
Heres some simple math. You have a 48 port switch. If even 5 clients want to stream stuff on switch 2… Perhaps copying an iso from the server, 2 people want youtube HD, someone else wants to run some backup, how much bandwith is left for the people on switch 4 and 6 combined ?
If you can find SPF+ modules that will do 1G you could use the US-16-XG. At that point find the extra $100 per switch and use 10G.
As far as long runs go, SM transceivers generally have a range of 10km. So the technology will not be a limit for you.
1 Like
I think what we needed is the Unifi Switch 48… man that was cost effective… I wished I’ve waited for the Unifi products…
Well if you haven’t ordered anything… or if you can move over, do it !! The cost of everything else would be pretty much similar to what you were planning.
If you already ordered them you could look at the return policy…
Sad part…Switches are already here and we can’t return them.
That sucks. The best option is to run as much fiber as you can to get two links from the core to each switch.
1 Like
Heres something else you can consider, never tried, but a good project, look at Ubiquiti AirFiber, it should higher latencies since its over air, maybe as a failover for the furthest points which have no redundancy.
1 Like
That could be an option but we might not be able to use LACP for Switches 3,4,5,6 that will be directly connected using fiber to the core switch…
Thanks… We tried the M5 powerbeam and get’s the job done ok… Consider the Ubiquiti AirFiber too.
You should be able to lag two fiber ports together just like two copper, if you can get them to run the fiber.
1 Like
yes sir thanks… My problem is that the switch only has 4sfp ports… So the option the we will try is just do a LAG of 2bundle of fiber and connect them to Switch 1 and Switch 2…
Well as long as your connecting it to the root switch, that’s definitely going to help your throughput
You could always extend the number of sfp by getting the mikrotik switch for 100 something, and run a 10gbe from that to your main switch and have the additional ports on that
1 Like
I get what you’re saying. If they haven’t pulled the fiber yet, get them to have two runs from the core location to each switch for some point in the future as it most likely will be needed.
1 Like
Yeah I think that’s the best option… Even though we can’t do direct fiber link from each switch to the core… Thanks!