I'm seeing some unexpected results with my MikroTik hEX S router.
Hope the experts here can help me understand what the problem might be.
Situation:
I recently upgraded our broadband link from ADSL2 to FTTP
FTTP upgrades WAN link download speed from 50Mbps to 900Mbps
I setup an iperf3 test to a privately-hosted AWS server
With a test device connected directly to the router, the Service Provider's router consistently provided the 900 - 930 Mbps download speed over many iperf3 tests (tested every 5 mins for 1 day)
Problem:
When I switched to the MikroTik hEX S router, the same test provided inconsistent results and with a lower average download speed.
See below for an example test result. Note the variability: 789 - 927 Mbps.
Compared with the service provider's router, the speeds are more variable and the average is generally 100 - 150 Mbps lower.
Repeating the tests over a week gave similar results.
Things I've tried - with no improvement:
Swapping the router devices & retesting.
Resetting MikroTik router configuration back to factory, & reconfiguring.
Upgrading MikroTik router to current version 6 (LTS)
Upgrading MikroTik router to current version 7 (stable)
Monitoring MikroTik router CPU when under sustained maximum download (CPU-1 & CPU-2 remained <= 85%)
Checking MikroTik WAN interface performance, by using Service Provider router as internet gateway. Reconfiguring MikroTik router as a DHCP Client connected to Service Provider router. Running iperf3 via the MikroTik router WAN interface to an iperf3 server connected to Service Provider router's switch. (Observed a stable 930+ Mbps over the WAN interface. Which indicates MikroTik device can perform as required with NAT, Firewall etc all active.)
Example iperf3 test results for MikroTik device, when acting as WAN gateway:
I pasted in a block of text from a Google Doc. Then selected the whole block. Then pressed </> for the whole block. The title lines got messed up. So I went back again and applied </> in smaller blocks. But not line-by-line. Weird, I agree.
Aside from running speed tests,
then, when you actually browse, not all sites offer more, just if you ask more...
Each has its limits... and maybe the total traffic you generate won't even fill the RouterBOARD...
I agree with @rextended, your results are generally in line with what you can expect. I know this is not what you're looking for, but I would seriously consider leaving it alone (whatever you gain may not be worth it in either time or money.)
Basically all newer Mikrotiks have better cpus, so they will give better results. E.g. the ax3 would be an ideal candidate.
Bear in mind that Mikrotiks handle pppoe entirely in software (even in fastpath) but many isp provided routers use hw offloading, which is bound to be more deterministic. Also, the isp router may be aware of your isp's support for increased mtus, which could also explain some of the discrepancy.
I only sporadically deal with pppoe, so I'm not the best person to ask.
You should find plenty of docs/tutorials on this. It's usually referred to as rfc4638 support. Of course the first step is finding out if your isp supports it or not, which is often easiest by trying it.