Hello,
I have a client that will avail of a 35Gbps line from us with BGP. Is simple queue still the way to go with this?
I’m planning to patch them into our core switch but then what? I’ve only ever tried shaping up to 1Gbps. Our core does not have connection tracking.
How do we shape these kinds of traffic levels? I’m thinking there must be a way to do this on the hardware level. The interconnect between me and client is 100Gbps.
You don’t have an aggregation-layer where you terminate such (very high-speed) customers ? Straight onto your core switch doesn’t sound like a very good plan to me.
I don’t know what CPU-power you have in your coreswitch-device, but perhaps you can iterate from 1Gbps shaping profiles to see howmuch cpu-time is going to that and make a guesstimate what will happen at 35Gbps
Semantics, I guess, regarding core/aggregation/etc. My main point is that all my equipment is MikroTik.
Yes, I haven’t terminated such a customer before. My previous record are 1G clients, which I do with simple queues.
My main question is are simple queues still the way, or are such speeds limited using another way?
I can’t speak out of experience, but I see no reason why simple-queues for the sake or shaping would work differently on 35Gbps vs 1Gbps.
What you need to try to find out is the impact on cpu & memory I guess. Shaping on IP-level like this using (simple) queues on 35Gbps is something you want to understand before implementing.
Reading things like
http://forum.mikrotik.com/t/one-simple-queue-consume-all-cpu/112056/1
http://forum.mikrotik.com/t/ccr1072-100-cpu-after-pcq/145606/1
“”
MikroTik made some changes to simple queues years ago that improved performance and gave them multi-core support (each simple queue item (or a group of them?) has a core for itself). So the solution was to use simple queues instead of queue trees. Oh boy and do I hate simple queues…well, turns out that adding about ~37500 simple queues halfbricks the router (reset needed) and the single core limit of ~1370Mbit/s still stands on the single core performance, so a single simple queue item can not limit more then the ~1370mbit/s on the ccr1036. And you need to have more then 32 simple queues for optimal load balancing on the cores, works ok with less tho.
“”
35Gbps must be done at hardware level, like on a switchport-level rate-limit.