New DSLReports speedtester
 

Just FYI.

http://www.dslreports.com/speedtest

It's tweakable too which is cool.

Re: New DSLReports speedtester
 

The upstream test always timed out for me on my VM circuit.

EDIT: Download showed my 154 meg.

Re: New DSLReports speedtester
 

Seemed limited to 100Mb on my download but upload worked OK

Re: New DSLReports speedtester
 

That's a really nice test! I really like that one, much better than Speedtest.net. My only complaint is they don't have a nice simple image to share, best you get is this:

http://www.dslreports.com/speedtest/234317

Edit: Just to put this out there, testmy.net is decent as well, though it never ever shows me hitting my top speeds so I'm not sure if it's less reliable or just harsher about real world speeds:

https://www.cableforum.co.uk/images/...2015/04/33.png

Re: New DSLReports speedtester
 

Now showing full speed

http://www.dslreports.com/speedtest/234358

Re: New DSLReports speedtester
 

http://www.dslreports.com/speedtest/235764

Seems OK. Not sure what the "Your link: Idle/Full" meter is supposed to mean, keeps flipping between 0, 4, and 100.

Re: New DSLReports speedtester
 

I think it tries to calculate the line capacity based on highest transfer rate reached then tell you if you're hitting it or not.

It's in beta so I fully imagine it's buggy right now.

Re: New DSLReports speedtester
 

http://www.dslreports.com/speedtest/341298 - I'm sure speed-testers love me skewing their results.

That was Plusnet FTTC balanced with VM 152.

Re: New DSLReports speedtester
 

Looks like that dial has been changed to (or replaced by) a "Buffer bloat" dial. The buffer-bloat dial shows any increase in your latency during the test beyond idle latency, which I suspect was also used for the "line capacity" indicator.

Very good metric to display IMO, I wish more speedtest sites and diagnostics would use it - it's a very basic tool yet very important tool that most consumer sites seem to miss completely. It's practically the first check I do to determine if a link is saturated or not.

Re: New DSLReports speedtester
 

The log is also very good, and I note it uses a few different servers depending on the number of streams in use.



---------- Post added at 18:43 ---------- Previous post was at 18:16 ----------

They do now!

https://www.cableforum.co.uk/images/local/2015/04/6.png

Re: New DSLReports speedtester
 

Ohhh snazzy!

https://www.cableforum.co.uk/images/local/2015/04/5.png

Re: New DSLReports speedtester
 

Looks like the upload test isn't working anymore for me :(

https://www.cableforum.co.uk/images/local/2015/04/3.png

Also it seems to use 32 streams on the download nomatter what, but the upload varies between 1 on my home connection and 24 on the above connection which oddly failed.

---------- Post added at 19:51 ---------- Previous post was at 19:45 ----------

Eh, looks like it was just Chrome on that machine, it works using IE:

https://www.cableforum.co.uk/images/local/2015/04/4.png

Interestingly, the bufferbloat measurement doesn't show up in IE but does in Chrome, but either way I'm impressed it works seemingly without flash, java, or any other junk. The ping measurement is also way off in IE.

Currently a complete lack of UK locations though sadly. Looks like they use a combination of servers from Google, Amazon AWS, and NFOservers.com

Re: New DSLReports speedtester
 

I think the download/upload streams are determined by which you pick - cable is 24/8.

Re: New DSLReports speedtester
 

Must be more to it than that, I picked "fibre" on every machine and it used 1 upstream on one machine and 24 upstreams on another.

Re: New DSLReports speedtester
 

I think it does a quick test with a small amount of data and from there decides on a number of streams.

Re: New DSLReports speedtester
 

https://www.cableforum.co.uk/images/local/2015/04/2.png

Sigh ... :D

Re: New DSLReports speedtester
 

Yup, I think you're both right. A few more tests show not a lot of difference on my connection.

Re: New DSLReports speedtester
 

Actually I'm not sure on that one now.

I just used the DSL setting and received 8 streams each way.

Oh well, still maxed out the connection so all good.

Re: New DSLReports speedtester
 

Well there's always the manual setting tab

Re: New DSLReports speedtester
 

Thought you were on FTTC quas?

Re: New DSLReports speedtester
 

VM D1 10/0.5 2008-2009
VM D3 50/1.5 2009-2011
BT FTTC 80/20 2011-2014
BT ADSL 11/1 2014-2015
BT FTTC 80/20 Last week till present.

Re: New DSLReports speedtester
 

for reference

no QoS on router, using plusnet pppoe.

buffer bloat C-E it varied.

no QoS on router, using sky dhcp auth

buffer bloat usually B but sometimes C

egress QoS on router using plusnet buffer bloat usually A but sometimes B
same on sky buffer bloat always A

QoS on router for ingress and egress on both isp's the buffer bloat is either A or A+ every time.

Very good tester.

Generally for most activities downstream QoS probably isnt needed as a low amount of threads on downstream will generally increase latency by a moderate amount only, whilst on upstream even a single threaded upload can make latency sky rocket. However I observed e.g. on steam downloads using no QoS on ingress has a pretty nasty effect causing packet loss for all other traffic. Steam bombard's the connection with 20+ connections during downloading and it seems the buffer gets completely swamped, doing QoS with 2mbit topped of my download speed stops the packet loss.

Other observations.

When restricting upload speeds and reducing upload buffers generally uploading is still able to flatline at the configured limit even single threaded, however doing the same on downstream will cause the download speeds to bounce around a bit at a bit under 100% utilisation (sometimes less depending on the congestion control etc.) which some people may not find acceptable (I expect if is any isp congestion speeds would plummet as the connection with reduced buffers will be much more "polite"), however multithread downloads can still typically flatline at the configured speed (e.g. steam).

Re: New DSLReports speedtester
 

Your buffer settings would have no effect on ISP congestion because the buffering occurs within the ISP. Your window and ACK settings may have an effect though.

Re: New DSLReports speedtester
 

the isp wont buffer if you move the bottleneck.

here is an example of ingress QoS.

Re: New DSLReports speedtester
 

Why would you deliberately bottleneck yourself if the ISP is congested?

Re: New DSLReports speedtester
 

Just tried this out, Wireless, Windows 8.1 64bit, IE11, no changes to Test parameters.
Granted sitting about 20 feet from Access Point and looks like it used my Infinity circuit for uploading.

https://www.cableforum.co.uk/images/local/2015/07/8.png

Re: New DSLReports speedtester
 

if the isp is congested the QoS goes out the window as then is no consistent line capacity.

If you were suggesting the line would be not worse affected under congestion, then yes maybe. But depends on the congestion I think, if the congestion hits after the buffer has grown then the larger buffers will probably give more throughput before they shrink. Thats my reckoning anyway.

In short router side QoS is pointless and a bad idea if there is frequent isp congestion.

Re: New DSLReports speedtester
 

I cannot figure out how to get the images people are using above, is that a different tester ?

Re: New DSLReports speedtester
 

https://www.cableforum.co.uk/images/local/2015/07/5.png

https://www.cableforum.co.uk/images/local/2015/07/6.png

The two tests are similar enough for me to stick with ookla as I like the way my results are illustrated. I guess the technical information contained in the DSLReports is more useful to members who understand what it means.

Re: New DSLReports speedtester
 

So I figured out how to get the graphic to display, and ran another test.

https://www.cableforum.co.uk/images/local/2015/07/4.png

My download speed is limited by the fact my router(s) only run 100M (Duplex) ports.

The upload is VirginSky. :)

Re: New DSLReports speedtester
 

https://www.cableforum.co.uk/images/...2015/09/27.png

http://www.dslreports.com/speedtest/1316591

Re: New DSLReports speedtester
 

Are you missing 50mbit or are you a lucky early upgrader?

Re: New DSLReports speedtester
 

https://www.cableforum.co.uk/images/...2015/09/28.png

Damn students coming back and hogging all mah bandwidths.

Surprised a flat 500/800 isn't "A+" for speed or quality. My e-peen has been well and truly flaccided by that.

Re: New DSLReports speedtester
 

Wow that is some serious downstream bufferbloat.

with router QoS - http://www.dslreports.com/speedtest/1079282

without router QoS - http://www.dslreports.com/speedtest/1289338

Re: New DSLReports speedtester
 

159.2 down and 12.39 up on mine
Bufferbloat C
Quality A
Speed B

Re: New DSLReports speedtester
 

I suspect that's just a bugged result, the bufferbloat readings on that can be all over the place:

https://www.cableforum.co.uk/images/...2015/09/24.png

Not a bad speed though, considering I'm only paying for 250Mbps.

Re: New DSLReports speedtester
 

Not actually sure what bufferbloat is but here is a test from a minute ago

https://www.cableforum.co.uk/images/...2015/09/25.png

Another one a minute later

https://www.cableforum.co.uk/images/...2015/09/26.png

Re: New DSLReports speedtester
 

I think their test is reliable as I can get consistent results and manipulate them in certain ways, remember VM are known for buffer bloat.

Of course since it is a browser based test the local testing environment can also affect results such as the browser been under pressure for resources, or perhaps a cpu hitting a bottleneck.

Tony bufferbloat is basically a big buffer, the test is basically testing how your latency compares between idle and when downloading/uploading, things like pings get delayed waiting when is lots of buffers full of waiting packets.

Whilst qasi may have those speeds which are probably only useful for warez, I wouldnt be too happy with latency measuring in the seconds whilst uploading.

Also usually downstream has minimal effect on latency providing the downloading isnt too agressive. Thats why I mentioned the VM test as it unusually had very high latency during the test. Poor bufferbloat on a upload test is not unusual without any QoS.

Re: New DSLReports speedtester
 

Re: New DSLReports speedtester
 

Thank you for the info. Would you say my line needed looking at with a F,D,F grade from the 3 tests?

Here is my TBB monitor if needed.

http://www.thinkbroadband.com/ping/s...11-09-2015.png

Re: New DSLReports speedtester
 

Tested just now ;

https://www.cableforum.co.uk/images/...2015/09/23.png

Re: New DSLReports speedtester
 

tony your graph looks normal for someone on cable, there is jitter due to how cable works.

It would seem the very high latency on your test is either down to the the router you are using or VM having a large buffer their side. Whether or not is a problem depends if having higher latency during downloads bothers you or not, I would expect if you had a download going at those speeds such as a steam download and you were web browsing at the same time with a 2000ms latency the browsing would feel very slow.

Re: New DSLReports speedtester
 

https://www.cableforum.co.uk/images/...2015/09/22.png

Was actually really happy with mine. I am lucky to never have issues with oversubscribing in my area. I do use a firewall so was able to bring that bufferbloat down.


Rexz

Re: New DSLReports speedtester
 

I get anywhere from A to F repeating the test several times on the same symmetric gigabit line. My transmit buffer is set to 20ms yet the the test simultaneously measures upload buffer to be 150ms +/-75ms (very inaccurate in itself) and 540ms +/- 250ms, aka all over the place. I wouldn't trust anything that thinks my latency is 75m and 750ms at the same time.

On another note VM may be known for bufferbloat but I have never seen downstream buffers of anywhere near 1800ms on VM. The highest I've seen is 150-600ms, and the latter only in extreme conditions.

The latency test also swings pretty wildly, probably just very inaccurate timing used in some browsers.
His latency is 1.8 seconds while downloading and 0.2 seconds while uploading.

Absolutely not.

---------- Post added at 17:20 ---------- Previous post was at 16:39 ----------

[Edit]

Yeah, definitely not what I would call consistent results.

IE ---

Test 1: FAIL - Upload too fast


Test 2: ERROR - Log too long
https://www.cableforum.co.uk/images/...2015/09/16.png


Test 3: FAIL - Upload too fast


Test 4:
https://www.cableforum.co.uk/images/...2015/09/17.png


Chrome ---

Test 1:
https://www.cableforum.co.uk/images/...2015/09/18.png


Test 2: ERROR - Your connection appears to be faster than 1 gig (with 830Mbps shown on the test status?!)


Test 3:
https://www.cableforum.co.uk/images/...2015/09/19.png

And yesterday:
https://www.cableforum.co.uk/images/...2015/09/20.png
I mean come on, 1020Mbps? Yeah, right...

Re: New DSLReports speedtester
 

something up with your system qas, is consistent here.

Also its a test probably aimed at consumer's not university connections? why are you using your work connection?

Re: New DSLReports speedtester
 

Yes, clearly it's my system because other speedtests work fine on my system...

There's specifically multiple options for gigabit, fibre, and work/corporate connections though as you can clearly see from the ISP box I am not using a university or work connection. But if it wasn't aimed at all connections (not just consumer) there shouldn't be a 32-stream option or big buttons for "Corporate" "Edu" "Fibre" "Gigabit" and "10Gigabit"

tl;dr: It's a beta. It can't be expected to be consistent and reliable all the time.

Re: New DSLReports speedtester
 

Your connection is probably faster than theirs :)

Re: New DSLReports speedtester
 

They use 4-6 servers including between 4-8 connections to each server, and that's for the non-10G test. 32 connections across 6 1G+ servers should certainly be enough to max out any connection I have. I dunno what they use for the 10G test, that's members only.

Given 4 streams is enough (on my line) to hit 900M+ I find 32-streams is rather excessive, but should have no problem with "my connection is probably faster than theirs".

Re: New DSLReports speedtester
 

Sky Fibre Pro
https://www.cableforum.co.uk/images/...2015/09/14.png

http://www.dslreports.com/speedtest/1350371

Re: New DSLReports speedtester
 

Depends how their load balancing those connections and there's a bit more to it than total bandwidth available per stream. I'd like some more technical details of how they are generating their test data.

Re: New DSLReports speedtester
 

The whole thing is HTML5 - just look at the source yourself?

Re: New DSLReports speedtester
 

Me and HTML5/CSS/Javascript don't get along :(

Re: New DSLReports speedtester
 

Well then you may have difficulty "getting along" with the technical details of how the test works - i.e. HTML5 websockets.

Re: New DSLReports speedtester
 

More interested in the technical aspects of packet size, UDP vs TCP, etc.

Re: New DSLReports speedtester
 

That's defined in the HTML5 websocket.

Re: New DSLReports speedtester
 

And the nature and configuration of the NIC's accepting those sockets? There's a lot of variables to consider.

Re: New DSLReports speedtester
 

Given there's dozens of servers across just as many countries, the NIC configuration is going to be different between all of them.

Re: New DSLReports speedtester
 

That doesn't sound like it's going to produce reliable, consistent results then. NIC configuration can make a hell of a difference.

Re: New DSLReports speedtester
 

Hence the brute force method, pile in 32 streams from 6 different servers.

For reference, Speedtest.net have no control over NIC parameters either, or even what kind of web server or connection is used, let alone the settings within. All servers there are just set up whatever way the sponsor/provider feels like.

Re: New DSLReports speedtester
 

Re: New DSLReports speedtester
 

its not supposed to give you always a perfect result, then it wouldnt be much of a tester.

if a NIC configuration causes issues then I would expect a good tester to show those issues.

I see these sort of posts time and time again, the tbb tester gets complaints also as its apparently broken due to showing issues other testers dont (due to its single threaded tests).

Re: New DSLReports speedtester
 

Finally get a symmetric connection, shame the d/l is only 2% of what I am paying for...

Re: New DSLReports speedtester
 

Nice. Is that CPE connected over USB or are they actually handing out 10GBASE-T gear?

Re: New DSLReports speedtester
 

10Gb.

Re: New DSLReports speedtester
 

I assume it's a SMB product and not a consumer service then?

Re: New DSLReports speedtester
 

It's $300 a month. SME or loaded home user.

Be ideal for a house share full of nerds, though.

Re: New DSLReports speedtester
 

Without a doubt. Especially for Offensive Security.

Re: New DSLReports speedtester
 

Even in a house full of nerds, I'd be surprised to find a 10GbE switched network. And my house is pretty high up there when it comes to the nerdiness of the nerds.

---------- Post added at 19:15 ---------- Previous post was at 19:13 ----------

That said, a GbE switch with 10Gb uplink wouldn't be quite as far-fetched, though you wouldn't be able to wave the e-peen with 2Gb+ speed tests without some bonding.

---------- Post added at 19:17 ---------- Previous post was at 19:15 ----------

$300 is pretty cheap though for 2Gb symmetric. You'd be hard pressed to even get a 300Mbps service (via BT FTTPoD) here for that much, heck you could find yourself paying $100+ for an 80Mb service.

Re: New DSLReports speedtester
 

2.5Gb should be knocking around in numbers in the not too distant, both as an uplink for other GigE ports and no doubt as access ports with 10GbE uplink. Broadcom 1GbE/2.5GbE SoC. Marvell 1/2.5/5/10/40GbE

Re: New DSLReports speedtester
 

Why bother with 2.5G? We already have the technology for 10G over ethernet, why not just focus on bringing the price of that down?

Re: New DSLReports speedtester
 

Same reason they bother with 40GbE while we already have the technology for 100GbE

Re: New DSLReports speedtester
 

I refer you to this presentation from Greg Hankins at Alcatel-Lucent delivered yesterday.

Re: New DSLReports speedtester
 

Incremental upgrades are more profitable for the hardware companies. So the answer is capitalism.

Re: New DSLReports speedtester
 

An 8 port gigabit switch is £20. An 8 port 10Gb switch is £600. A gigabit NIC from a recognised brand is £6. The cheapest, unknown brand 10Gb NIC I could find is £200. The answer is price versus performance making sense.

The sweet spot is about 3-4 times the price for 10 times the bandwidth. Right now 10GBase-T isn't even close at 30x.

As the presentation indicates there are genuine use cases in offices and, in time, homes, for 2.5Gb and 5Gb. The idea that it's because incremental upgrades are more profitable seems strange given it's not incremental, 10Gb has already been released, it's just too expensive for the applications that may use it.

EDIT: I glossed entirely over something else quite obvious, too. The vast majority of cabling is Cat 5. 10Gbase-T doesn't run so well on most Cat 5, so something that will allow upgraded bandwidth without requiring changing out of all cables in offices is very desirable.

Re: New DSLReports speedtester
 

That was a really interesting link and I read it all, but I'm still a little unconvinced. Their mean reasoning for 2.5G and 5G to exist is that wireless speeds have surpassed gigabit - and it's a worthy point. However, on the same slide they state this, they show that wireless speeds are going to go beyond 5Gbit in less than 2 years and faster still in 3 years.

https://www.cableforum.co.uk/images/...2015/09/11.png

It still seems like a bit of a pointless stop-gap between 10GbE? Then again, the cabling aspect is a worthy one and it's certainly better to be able to negotiate to 5Gbit if 10Gbit isn't possible on that particular cable. I suppose that's the real difference, having something inbetween for when your cabling isn't good enough. Still, it'd be nice if this just meant that devices started shipping with 10GbE capable ports that could just negotiate down, rather than trickling devices that do 2.5GbE and 5GbE.

Re: New DSLReports speedtester
 

10gbit is out there but not aimed at the consumer market. As a consumer product 2.5gbit first makes sense.

Look at across the entire hardware space and you see the practice is incremental upgrades, from smartphones to computer components.

No way do hardware vendors release the best they have to offer, they always keep something in the pocket to get your money another day.

As an example intel have been making 16 core cpu's for years, yet we only got their first consumer 8 core cpu earlier this year.

Re: New DSLReports speedtester
 

Hardware vendors don't release the best they've got because it's too expensive.

10GBase-T has been around since 2006. If there were demand for it there would be more units produced and it would be more common.

You're complaining about 10GBase-T being an unnecessary incremental upgrade. I'm just pointing out that it is not. Something is needed to fill the gap as 10GBase-T hasn't come close to the levels of usage it was expected to originally.

Of course hardware vendors don't release their absolute state of the art as consumer products except in rare occasions like GPUs where some nutters will pay nearly £1k for a graphics card. It's. Too. Expensive.

Intel have been making 16 core CPUs, yes. How many people run applications that actually make use of 16 cores? I have a machine that has 16 cores, 2 x 8 core Xeon. It's my home lab machine and runs ESXi. I rarely have it switched on as it eats electricity like a beast and blasts out so much heat it raises the temperature noticeably in the floor above. I would suggest that's not a common usage scenario, so they have been targeted at businesses.

The big power whores for CPUs are gamers, for whom most games don't make much use of multithreading and fewer, faster cores are preferable, and things like video transcoding, which is something that lends itself well to multithreading but will be done on enterprise kit.

It's the same argument as with broadband. People don't want to pay for the latest and greatest for the most part, they just want good enough. The vast majority of VM's customer base continue to take the lowest sold product.

Re: New DSLReports speedtester
 

I would prefer 10gbit over 2.5gbit, but I am just saying how business looks at it.

Re: New DSLReports speedtester
 

Then you're in luck.

Re: New DSLReports speedtester
 

It's a lot pricier than it should be :(

Re: New DSLReports speedtester
 

A 160Gb switching fabric doesn't come cheap.

Re: New DSLReports speedtester
 

I'm not sure how seriously I can take an "Enterprise" vendor trumpeting 5Gbps cat videos as a use case for their technologies.

---------- Post added at 19:40 ---------- Previous post was at 19:33 ----------

Probably because there's little benefit to it on edge devices partly because they never expected hard drives to be stuck at 50% more STR than they had 10 years ago. That, and the fact that it's not going to go anywhere near a laptop or low-end home PC until it starts using less idle power than the whole PC.

On the other hand, I don't know a single datacentre core that doesn't run on 10GbE already. Everywhere I know that actually needs 10GbE has it, with many moving beyond.

Still, as I mentioned before I suspect the pricing and take-up may change significantly over the next couple of years thanks to certain very large, powerful companies bundling it into mass-market consumer gear at below-cost prices.

And you call me a nutter for my 24-bay NAS with a single 4-core Xeon? :dozey:

P.S. 22-core Broadwell-EP is due out in the next month or two.

---------- Post added at 19:45 ---------- Previous post was at 19:40 ----------

Wanna guess how much one of these costs?

http://www.cisco.com/c/en/us/product...-2t/index.html

Re: New DSLReports speedtester
 

Just FYI the kit they supply is one of these.

Yes really.

Re: New DSLReports speedtester
 

Included in the $300 a month or extra?

I'd normally say that'd definitely make it a SME product and not "rich home user" but that said, it's fanless. Fanless.

Re: New DSLReports speedtester
 

It's included in the price.

It's environmentally hardened kit, it's intended more for use as an access router than taking a 10Gb feed for a property.

I've a suspicion Comcast know they won't sell many and are reusing surplus metro net equipment as CPE :)

Re: New DSLReports speedtester
 

That much is obvious from the 16 TDM ports and the "next-generation access routers" line in the description :p:

Should be perfectly sufficient for your 600Mbps+ dual-WAN requirement though :D

Re: New DSLReports speedtester
 

Speedtest.net has put up a beta test.

http://beta.speedtest.net/result/4701464425

Re: New DSLReports speedtester
 

Flash based, but seems to work ok ;

https://www.cableforum.co.uk/images/local/2015/09/2.png

http://www.speedtest.net/my-result/4701931336

Re: New DSLReports speedtester
 

Working just fine.

https://www.cableforum.co.uk/images/local/2015/09/1.png

Re: New DSLReports speedtester
 

Yes, we mentioned it several months ago.

Re: New DSLReports speedtester
 

working fine ....

https://www.cableforum.co.uk/images/...2015/10/14.png

Re: New DSLReports speedtester
 

On the new 200Mb VIVID
Doesn't even burst the 150Mb limit :td:

http://www.thinkbroadband.com/speedt...97655-mini.png

Re: New DSLReports speedtester
 

https://www.cableforum.co.uk/images/...2015/11/31.png


https://www.cableforum.co.uk/images/...2015/11/32.png


The U Switch tester seems pretty accurate too, however the naff results presentation in 'Street Stats' is a complete let down to anyone visiting the site to get some comparative data. In fact it borders on the negligent for a serious comparison site.

http://www.uswitch.com/broadband/speedtest/

Re: New DSLReports speedtester
 

https://www.cableforum.co.uk/images/...2015/11/30.png

No bufferbloat C rating here ;)

Re: New DSLReports speedtester
 

Me neither.

https://www.cableforum.co.uk/images/...2015/11/29.png

(Talk about one-way internet...)

Re: New DSLReports speedtester
 

Finally looking decent:

https://www.cableforum.co.uk/images/...2016/02/12.png

Re: New DSLReports speedtester
 

Re: New DSLReports speedtester
 

https://www.cableforum.co.uk/images/local/2016/02/9.png

https://www.cableforum.co.uk/images/...2016/02/10.png

http://www.thinkbroadband.com/speedt...12355-mini.png

One after the other, not very consistent....

Re: New DSLReports speedtester
 

First time trying the DSLReports speedtest

https://www.cableforum.co.uk/images/...2016/03/21.png

Re: New DSLReports speedtester
 

Was trying to post yesterday, but internet was very slow yesterday down to 20mb, recovered after midnight.

https://www.cableforum.co.uk/images/local/2016/06/2.png