For the benefit of others, @linuxuser and I had this conversation via pm, but worth sharing here…
Simple... the 9.4Mbps speedtest was performed with the router itself as the source or sink of the packets (iperf directly installed on the router). Routers don't have very powerful general purpose CPUs... the power they have is in the specialized routing hardware.
the point remains that iperf on the router itself is not representitive of the speed at which the router can move traffic via routing.
Case in point... The Ubiquiti Unifi Security Gateway (USG) is an older 1Gbit router. It can legitimately route at 1Gbps. But, if you run a speed test on the device itself, it will cap out at 250Mbps.
The magic is in designing hardware accelerators and offloading the routing to a specialized routing engine.
May wildly fluctuation sync rats often denote a link with some noise issues... I would probably first try to stabilize the DSL-situation before moving deeper into the home network.
Because IIUC you measure via WiFi and WiFI rates can also fluctuate wildly in relativ short amounts of time?
Should not matter since we are not trying to get json output, however I have not tried any variant of that version, but I assume it will just work as expected.
Well, you should find well connected servers close by (either within your ISPs network if you want to test the best case, or in another AS if you also want to test your ISP's peering/transit). The throughput of TCP is inversely proportional to the RTT (at least under competition) so aim for short RTTs, unless you explicitly are interested in throughput over long RTT paths, but even then start with short RTT paths.
That is not how it should be and not how I remember this, however you realize that you need to use two different invocations for testing upload and Download speed with iperf3? From man iperf3:
-R, --reverse
reverse the direction of a test, so that the server sends data to the client
--bidir
test in both directions (normal and reverse), with both the client and server sending and receiving data simultaneously
so unless you specify either -R or --bidir you will indeed only measure your upload. So what invocation did you use?
Yes, all of this seems as expected when iperf3-ing to a remote server with the default option. Maybe try adding -R?
+1; that is not very intuitive, but backed up by experience... for very beefy hardware (e.g. XEON class) this is a non-issue, but typical cheap home routers are extremely cost optimized and tend not to come with CPU cycles to spare.
Hrm.... that is indeed what a number of router makers aim for... and all the way at the top end of the market this is what big iron routers also do, move as much of the processing into hardware as possible. And that makes some sense in such environments. However for home routers having a reasonably beefy CPU will allow quite a bit of networking in software and can give a router a longer lease on life (as features can be added as software later on*).
*) There was a recent example where a rural ISP decided to take their (abandoned by the manufacturer) home routers and upgraded them to OpenWrt immediately allowing/offering sqm/cake; I am not saying this is impossible for accelerators but it is certainly easier in software.
Most all-in-one or purpose built router devices have routing specific hardware acceleration, AFAIK. Small devices like the ER-X (MSRP was ~$50 USD) rely on it... if you disable the HW acceleration, it drops down from 1Gbps routing to around 250Mbps max. The SoC in many of these devices are basically routing engines combined with a low-ish power general purpose CPU to control the whole system and provide software features. These chips are considerably more efficient and cheaper than a larger general purpose CPU that would have the bandwidth to route at line-rate.
Also of note -- HW acceleration (and the lack of appropriate open source support for some of these routing engines) is part of the reason that some devices have relatively poor performance when running on OpenWrt (or other FOSS firmware) when compared to the vendor's firmware.
Why should it be the other way around? The Internet is mostly used to download files from the Internet, but not to send files to the Internet. There is not enough bandwidth for all users, so the download speed is lower than advertised.
Because if I measure the speed to a server in my own city, then I am essentially measuring the speed not of the Internet, but of the local city network. But I don’t download something from the city’s servers, but from a server much further away. Within the city limits, the test can show all 100 Mbit, but two people will fall into the neighboring city.
I understand all this, but it doesn’t suit me. I don't need high speed on the city's local network. If my speed is 100 Mbit in the city and 25 Mbit outside the city, is this normal? According to the tariff, the provider promises me 50 Mbit, but there is a tariff for both 100 Mbit and 1000 Mbit. But if he can’t even provide me with 50, why pay more? Of course, within the city everything will be quick.
My provider will distribute the Internet only in my city. He lets me into the external Internet through a backbone provider. Let's assume that I bought an Internet tariff for 1000 Mbit, and in the city I have 1000 Mbit. But the backbone provider has an Internet speed of only 500 Mbit. It turns out that no matter how hard I try, I can’t access the external Internet at the speed of 1000 Mbit that the local provider promises
PC via cable
PC via Wi-Fi
Smartphone via Wi-Fi
Smartphone via Wi-Fi from PC
The Internet cable does not reach the PC. And it’s not convenient to move the PC every time you need to make speed text
But your local ISP cannot control the rest of the internet. The ISP is selling you a package that guarantees they will deliver the plan's speed to you via their network.
If I run a site on a slow server/ISP, you cannot hold the ISP accountable for the slow speeds you get from my site.
Or if you buy an item online and pay for overnight delivery but the seller doesn't actually ship it for 3 weeks, you cannot hold the shipping carrier responsible for the 3 week delay since they met their part of the deal. Your contract is with the ISP that is effectively the shipping carrier.
So regardless of if it suits you or not, that's the way it is... don't like it? Start your own ISP.
Looks like you're getting the 50Mbps you're paying for.
looks pretty clear that the problem is the wifi, wouldn't you say? Some of that may be resolved with a better AP, but physics plays a part, too, so it is possible you may not see improvements without other methods..
I hope it is clear to you that you are getting what you are paying for. Your wifi is clearly a real bottleneck. And your tests to distant servers, while interesting, are not guaranteed by your ISP (regardless if you like it or not... that's just the way it is, full stop.).
If a local provider is deceiving me and offering a speed that a higher-ranking provider cannot provide, then I should know this so as not to pay him extra money
Then what router should you buy to guarantee Internet speeds of 50-100 Mbit over the air? And so that WireGuard would work normally. But the Linksys EA8300 is not offered, it's too expensive..
You are seriously mistaken. However, if you feel that way, cancel your service and find another ISP. Be sure to read the fine print
Edit: here is a link to Comcast’s legal disclosure regarding network speeds (for readers who do not know, they are one of the largest residential ISPs in the US) - they make it extremely clear that the speed you actually get is very dependent on both your own equipment/conditions and also the servers and networks which host the content and services you wish to access. Look at point 3 for the full description.
Well, pretty much any well designed 802.11ac or 802.11ax router these days can provide that, but they are still subject to physics when it comes to distance, penetrating walls, congestion/noise/interference in the airwaves, etc.
My Unifi AP-AC-PRO gives me 400Mbps+ maybe 10 feet from the unit with 80MHz channel width on the 5GHz radio. That speed drops in half just a when I double the distance to ~20 feet in another room (thus also penetrating a wall). These are particularly well designed 802.11ac devices (although newer generations are better) and I would classify them as excellent APs... but you can see how much the environment matters (inverse square law + the permeability of the materials at 5GHz).
It was $129 USD, iirc. You can get the U6-lite which will outperform mine for just $99.
That is really not that expensive.
P.S. please check the link I added to an earlier post with Comcast’s terms of service. You should read it because it clearly identifies the delineation between their service speed tiers and the actual performance you may see in practice.
That's your point of view as a customer...but the ISPs, Internet exchange points and hoster may have a different idea.
It shouldn't be...at least not below for what you pay. And of course only for the part that is in control of your ISP.
I don't know if you've got a symmetrical or asymmetrical line. Here in western europe you'll mostly get asymmetrical lines where the (advertised) download speed is much higher than the upload speed (i.e. 16/1, 50/10, 100/40 and so on).
I was just puzzled about your upload speed with approx. 80 Mb/s compared to less than a half for download.
This is the first time I've heard about asymmetrical speed. I always have the same speed here. If the download speed is 100 Mbit, then the upload speed will be approximately the same
Well, this makes me a bit envious...
I've (well, we've) got a 100/40 VDSL line.
The faster alternatives are 250/50 via VDSL, 300/150, 500/250 and 1000/500 via fiber and 1000/50 on cable (docsis). There's nowhere an affordable offer for a higher speed symmetrical line.
But when you have a symmetrical line, you should get 50Mb/s in both directions. If you do a speed test with a local server (the nearer the better), do you get equal speeds in both directions or is the download still much slower than the upload?
A simple alternative: temporarily replace the router with a female-female ethernet coupler. Then your PC will be connected directly to the service. You can buy one on Amazon for under $10.
Not sure where OP lives, but I will assume hes living in the USA.
You can upgrade your router to a R7500v2 for about 50 dollars on ebay. Has a 1.4ghz ipq8065 SOC, dual band, and gigabit ethernet ports. also has USB 3.0 and esata if your into that sort of thing.
Asymmetrical speeds are common almost everywhere. You may downloading often, but there are others out there who also download and upload on average or upload more often than they download. Content creators are one good example.
Given his speedtests (reported location and RTTs) I am pretty sure he does not live with in the american continents but somewhere in or around ukraine.
