To get the best performance two simple rules must be taken into account:
• avoid packet fragmentation;
• there must be no packet reordering.
What measures should be taken to avoid packet reordering? I haven’t found any hints in the newsletter nor on the documentation page it points to. Does it also apply to the other multi-core products that support hardware-accelerated encryption?
That particular problem with hw acceleration applies only to CCRs.
For example queueing and packet buffering can lead to packet reordering, that could be avoided by decreasing buffers, changing queue algorithms for example to pfifo, etc.
That will make it more expensive, but the whole point of this LHG5 device is that it’s cheap.
For those willing to pay more there’s a DynaDish 5 available already.
So what?
I know, and tell about more powerfull device.
Half year ago we mounted bridge on 4 km with new DynaDish. But real real data throughput only ~280-300 Mbps (nv2).
Looking at new Rocket 5ac Prism by ubnt. 650 Mbps on 15 km - it’s easy.
I like Mikrotik and wait something more powerfull. LHG 5 realize good idea, but too simple realization.
Have you compared these devices in exactly the same conditions (literally - at exactly the same place and roughly the same time)? If not, your comparison is irrelevant.
Please note that I’m NOT saying that one device is better then another. Honestly, I have no idea which one is better. I’m just saying that your comparison may very likely be incorrect.
For rocket available 31 dBi antennas. DynaDysh is all-in-one.
But DynaDish has lower price.
Why Mikrotik dont do something like rocket, based on LHG design?
Could you, please, do not cite full post under cited post or two or three posts later ? It does not help reading the thread and makes the topic much longer than it should be. If you want emphasise some words just cite this part of the post.