[TUTORIAL] How to test the speed of your local network

[occamsrazor]

I have netgear 3700 router. usually i got around 9- 11MB/s when transferring files to my nas using 2.4ghz.
However some days ago i connected 5ghz band which my router is famous for not being so fast according to most reviewers. at 5ghz band i got around 15 MB/s.
So today i connect dual ethernet from my nas to my router with the setting active backup (fail over) i come up between 18MB/s-21MB/s at tops. Can drop below 18MB/s sometimes.

WNDR3700 is considered one of the fastest routers for 5Ghz wifi. 18mb/sec is a very good result I think.

[trevor294]

I tested by downloading the webpage into both a TS-219P and a TS-419P. I got results that were almost exactly the same for either QNAP device:
183000 Kb/sec downlink --> 151 Mb/sec wire speed (divide by 1024)
84000 Kb/sec uplink --> 82.0 Mb/sec

I also cross-checked the NAS's Bandwidth Transfer monitor at that time. It showed a peak of about 18MB/sec --> 144 Mb/sec (Multiply by . So, the numbers all make sense.

The numbers do seem a bit low for the network performance. Makes me want to do some stress testing with packet test equipment. Something tells me the Apache webserver is slowing things down a bit.

[booyip]

Haven't used this for a while and am now getting errors saying it has expired. Is it working for anyone still?

[Mibius777]

Is there a way to test Internet Speed of QNAP remotely using web interface?

[CylonCenturion]

A new version of mini speedtest is available.
Users should make an account first, and obtain the latest version from here

After that, unzip the file, open index-php.html, remove all text, and paste:

Code: Select all

<!-- BEGIN SPEED TEST - DO NOT ALTER BELOW-->
<script type="text/javascript" src="http://synergy-uk.com/speedtest/swfobject.js?v=2.2"></script>
	  <div id="mini-demo">
		  Speedtest.net Mini requires at least version 8 of Flash. Please <a href="http://get.adobe.com/flashplayer/">update your client</a>.
	  </div><!--/mini-demo-->
	<script type="text/javascript">
	  var flashvars = {
			upload_extension: "php"
		};
		var params = {
			wmode: "transparent",
			quality: "high",
			menu: "false",
			allowScriptAccess: "always"
		};
		var attributes = {};
		swfobject.embedSWF("http://synergy-uk.com/speedtest.swf?v=2.1.8", "mini-demo", "350", "200", "9.0.0", "speedtest/expressInstall.swf", flashvars, params, attributes);
	</script>
<!-- END SPEED TEST - DO NOT ALTER ABOVE -->

This is an updated script to tackle the ; update your client' error.
Please note that adobe flash needs te be installed.



Not bad on my TS-253 Pro 8Gb

[P3R]

Not bad on my TS-253 Pro 8Gb

Good that you're happy, but unless you're connected via WiFi or have some other known bottleneck, I'd think that 41.13/44.17 Mbits/sec is pretty bad.

[CylonCenturion]

Not bad on my TS-253 Pro 8Gb

Good that you're happy, but unless you're connected via WiFi or have some other known bottleneck, I'd think that 41.13/44.17 Mbits/sec is pretty bad.

When i download from usenet. I have a speed of 23 megabuyte

[P3R]

When i download from usenet.

What is downloading? The NAS?

I have a speed of 23 megabuyte...

What exactly does that mean? 23 MBytes/sec?

[CylonCenturion]

When i download from usenet.

What is downloading? The NAS?

I have a speed of 23 megabuyte...

What exactly does that mean? 23 MBytes/sec?

The ts253 is directly connected to a fiberspeed modem with a brand new cable.
Upon opening the nzbget webinterface and start dowloading something. The speed goes to 23 mb/sec.

[P3R]

The ts253 is directly connected to a fiberspeed modem with a brand new cable.

Let's for your sake hope there's a router in it as well.

Is the NAS connecting at gigabit speed (1 Gbit/sec=1000 Mbit/Sec)?

The speed goes to 23 mb/sec.

23 millibits/sec sounds very slow...

Without the use of proper units and prefixes it's absolutely meaningless to try and discuss performance. So do you mean 23 Mbit/sec or 23 MBytes/sec and in what way do you find that comparison related to the "...speed of your local network", that is the topic of this thread?

[CylonCenturion]

The ts253 is directly connected to a fiberspeed modem with a brand new cable.

Let's for your sake hope there's a router in it as well.

Is the NAS connecting at gigabit speed (1 Gbit/sec=1000 Mbit/Sec)?

The speed goes to 23 mb/sec.

23 millibits/sec sounds very slow...

Without the use of proper units and prefixes it's absolutely meaningless to try and discuss performance. So do you mean 23 Mbit/sec or 23 MBytes/sec and in what way do you find that comparison related to the "...speed of your local network", that is the topic of this thread?

have to go to the dockter.
Will post some screenshots around 1600h ( within neaarly one and a half our )

[CylonCenturion]

The ts253 is directly connected to a fiberspeed modem with a brand new cable.

Let's for your sake hope there's a router in it as well.

Is the NAS connecting at gigabit speed (1 Gbit/sec=1000 Mbit/Sec)?

The speed goes to 23 mb/sec.

23 millibits/sec sounds very slow...

Without the use of proper units and prefixes it's absolutely meaningless to try and discuss performance. So do you mean 23 Mbit/sec or 23 MBytes/sec and in what way do you find that comparison related to the "...speed of your local network", that is the topic of this thread?

so i installed google chrome in the hybrid station.
i opened in google chrome speedtest.net.

downspeed came up to 404.53 Mbps
upspeed came up to 397.36 Mbps

see also a pending download on TS-253.


so i simple count:

21 Gb in 15 minutes
21 Gb : 15 = 1.4 Gb per minute
1400 : 60 = 23 .4 Mb/sec ( the same speed according to my usenet abonnement of 200mbits )

a .bin file downloading via ssh on the NAS:


My simple conclusion is, is that this speedtest does not make any scense.
My NAS is in my opinion capable of streaming nearly 44 megabuyte per second

[P3R]

@CylonCenturion,
The use of capital letters in units and prefixes are not random or meaningless. What letters you use makes all the difference, so you'd better start paying attention to them. If you do, your numbers will start to make much more sense to you and in addition, you can start having meaningful discussions with other people on the subject.

In the datacommunication world (ISPs, networking equipment manufacturers, and performance measurement tools like speedtest.net) the unit used is almost exclusively bits/sec or more commonly nowadays with the prefix Mega (=1 000 000), Mbits/sec. For example your internet connection is probably marketed as up to 500 Mbits/sec (Mb/sec, Mb/s, Mbps also mean Mbits/sec).

In operating systems and applications the unit is most often Bytes/sec or more commonly nowadays with the prefix Mega, MBytes/sec. For example the numbers shown in both your latest screenshots here are in MBytes/sec (MB/sec, MB/s, MBps also mean MBytes/sec).

One byte consists of eight bits so 8 bits/sec = 1 Byte/sec. Some other commonly used numbers:
100 Mbits/sec = 12.5 MBytes/sec
1000 Mbits/sec = 1 Gbit/sec=125 MBytes/sec
10 MBytes/sec = 80 Mbits/sec
50 MBytes/sec = 400 Mbits/sec

Please notice that the difference between a "B" (Byte) and a "b" (bit) is a factor of 8. Therefore you can't use them interchangeably when reporting performance measurements!

The difference between the prefixes "M" (Mega = 1 000 000) and "m" (milli = 0.001) is so enormous that in this case we all understand that you can't really mean "milli" despite you writing that but if you want to be taken seriously, it makes sense to use the correct letters for prefixes as well.

In addition to the above you need to understand that:

• Different protocols have different performance characteristics and can give hugely different results over the same connection.
• Performance is always limited by the "weakest link". That bottleneck may not always be where you think, so you need to be careful to make sure that you know what it is that limits your performance before drawing conclusions from your testing.
• The topic of this thread is "...speed of your local network" and in your first post here I think that you really measured locally between the NAS and a computer, but I think it was with a slow WiFi connection as bottleneck, so you really only measured WiFi performance. Now your focus seems to have shifted to internet only, which is much more complicated as it's hard for a non-professional to understand where the bottlenecks are located when most of the network path is completely unknown.

[CylonCenturion]

@CylonCenturion,
The use of capital letters in units and prefixes are not random or meaningless. What letters you use makes all the difference, so you'd better start paying attention to them. If you do, your numbers will start to make much more sense to you and in addition, you can start having meaningful discussions with other people on the subject.

In the datacommunication world (ISPs, networking equipment manufacturers, and performance measurement tools like speedtest.net) the unit used is almost exclusively bits/sec or more commonly nowadays with the prefix Mega (=1 000 000), Mbits/sec. For example your internet connection is probably marketed as up to 500 Mbits/sec (Mb/sec, Mb/s, Mbps also mean Mbits/sec).

In operating systems and applications the unit is most often Bytes/sec or more commonly nowadays with the prefix Mega, MBytes/sec. For example the numbers shown in both your latest screenshots here are in MBytes/sec (MB/sec, MB/s, MBps also mean MBytes/sec).

One byte consists of eight bits so 8 bits/sec = 1 Byte/sec. Some other commonly used numbers:
100 Mbits/sec = 12.5 MBytes/sec
1000 Mbits/sec = 1 Gbit/sec=125 MBytes/sec
10 MBytes/sec = 80 Mbits/sec
50 MBytes/sec = 400 Mbits/sec

Please notice that the difference between a "B" (Byte) and a "b" (bit) is a factor of 8. Therefore you can't use them interchangeably when reporting performance measurements!

The difference between the prefixes "M" (Mega = 1 000 000) and "m" (milli = 0.001) is so enormous that in this case we all understand that you can't really mean "milli" despite you writing that but if you want to be taken seriously, it makes sense to use the correct letters for prefixes as well.

In addition to the above you need to understand that:

• Different protocols have different performance characteristics and can give hugely different results over the same connection.
• Performance is always limited by the "weakest link". That bottleneck may not always be where you think, so you need to be careful to make sure that you know what it is that limits your performance before drawing conclusions from your testing.
• The topic of this thread is "...speed of your local network" and in your first post here I think that you really measured locally between the NAS and a computer, but I think it was with a slow WiFi connection as bottleneck, so you really only measured WiFi performance. Now your focus seems to have shifted to internet only, which is much more complicated as it's hard for a non-professional to understand where the bottlenecks are located when most of the network path is completely unknown.

Thanks for your detailed answer, but i was completely ignored about b and B.

[gad]

Hi,
I just donwloaded the speedtest mini from http://www.speedtest.net/mini.php but after the installation it says "This test has expeired" Did I miss something?

Screen Shot 2015-12-28 at 00.38.16.png