I will be needing the extra SFP in the main stack, two MS-410-16 will do the trick as I won't be needing the extra SFP+. Thanks for all the feedback! ... View more

I did mention the 425 but am thinking it may be a little overkill for my application. What do you think about this... I am wondering if 2 MS410-16 would be the best option for the core layer. I already have 2 of each MX250 and MS350, 250, 225 in my MDF with most ports on the MS's already taken up. I would create three stacks from each model MS to serve the access layer with redundant and possibly aggregated 1g links to each MS410. The remaining 4 sfp+ will connect my branch locations.   The redundant MX is something I am still unsure how to implement while keeping both internet connections load balancing. Is it possible to run a spare MX with no active uplinks so in case of a failure I could quickly swap the cables from one MX to the other? I can't afford the extra circuit to keep internet access to both MXs which seems to be my only option if I am required to load balance my two existing internet connections. ... View more

I like this idea but should have mentioned that I would eventually like to see redundant or aggregated links to each branch since the fiber will already be there, this would be a minimum of 12sfp ports in the core. I might be wrong but I think a stack of 3 350s would still have less switching capacity than one 425. Do you think having the redundancy at the core while still having a single point of failure due to my MX would be my best route or would you say having the single MX defeats the purpose of the 350 stack?  ... View more

Thanks for the reply! I am far from a network engineer but have been tasked with configuring a campus to support 4 spoke locations from the internet connections terminating at our hub location. I may be overthinking the redundancy of it all but would really love to avoid any outages while staying within budget. I will attach a quick topology I drafted up to see if anyone notices any oversights on my part, the topologies of the spoke locations are not detailed or finalized. Keep in mind that we simply do not have the budget to implement redundant aggregation switches.   ... View more