Right, down to business. I’ve been following Dale Joyce’s post here: http://forum.mikrotik.com/t/nstreme-ptp-help/25457/1 closely, but unless I’ve missed something, no-one seems to have a definite answer to the following questions:
Can a high throughput link (70 Mbps or thereabouts), be achieved using RB333’s and R52’s?
If the answer is yes, how was it achieved and is there a detailed step-by-step guide?
If it is not possible using the above hardware, what would be the recommended hardware to use?
Yes, we get about 68Mbps using 333s and 433s in real life using turbo. Yes, you need clean spectrum to make it happen. Yes, you need to test through the RBs, not using them as end points because the sending machine requires lots of cycles.
This is not a trivial process, it needs to be engineered.
A suggestion, don’t use 333s, they are EOL. Use 433s instead. And test through them, not using them as end points.
approximately what thru put you get machine to machine at both ends as we have a similar requirement where in we have to transport 50mbps full-duplex to a Sub-ISP from the main Hub .
We managed to get the throughput required in the end, by replacing the RB333 with an RB433AH. We used R5H cards, but then swapped them back to the R52 cards and were still getting the higher throughput, so it must have been the RB333 boards that were the problem.
We achieved 70Mbps on a bandwidth test board to board, and 68Mbps laptop to laptop, transferring a 700MB file through FTP.
Thanks to George (ghmorris) for your help and advice. Mankomal, if you’d like an exact breakdown of equipment and configuration used, Dale will be posting it here: viewtopic.php?f=7&t=28651.
just a note - when doing wireless tests. you can also do them board-to-board because wireless speeds are not so great that it could be affected by CPU bottleneck. This is more important for ethernet tests, where the speeds are in hundreds, and hardware becomes more important
