Does Bittorrent Traffic Effect Speedtests?

[Rowan]

Hi,

I was just wondering if your ISP is doing Bittorrent Trafficking does that effect your Speedtest results?

Like ive been told the reason my upload speed could be unstable and slow is because im under trafficking, but when I do a Speedtest I get my full download and upload speed.

Thanks

:)

[PeteLockwood]

Should not do sweetest are http

[Rowan]

So doing a speedtest is no way to check if your definetly trafficked?

You could still get full speeds on the test but could still be trafficked.

[Ignitionnet]

OK let's put this to bed :)

http://broadband.mpi-sws.org/transparency/bttest.php

A tester for torrent shaping, enjoy!

[Rowan88]

Hi,

thanks for the replys.

I tried that test and it said it didnt look like I was being throttled which is good.

I also took my laptop with the same settings nothing changed to my friends house, and opened the port on there ASUS router and right now im getting all 4 things seeding at full speed possible :)

So it must be the ISP/modem...

My engineer is coming tomorrow hopefully and will sort it (if im lucky)

[Sephiroth]

I tested on VM with the following result:

Is your upload traffic rate limited?
There is no indication that your ISP rate limits your uploads.

Is your download traffic rate limited?
Your ISP appears to rate limit your downloads.
However, some of the measurements were affected by noise, which limits Glasnost ability to detect rate limiting.

Details:

Your ISP appears to rate limit your BitTorrent downloads. In our tests, downloads using control flows achieved up to 4880 Kbps while downloads using BitTorrent achieved up to 3568 Kbps.

I then tested on my O2 link and got this:

Results for your host (xx.yy.bethere.co.uk - zz):
Is your upload traffic rate limited?
There is no indication that your ISP rate limits your uploads.

Is your download traffic rate limited?
There is no indication that your ISP rate limits your downloads.

Any use to anyone?



There is no indication that your ISP rate limits downloads on port 6881 or 55534. In our tests, downloads on port 6881 achieved up to 4183 Kbps while downloads on port 55534 achieved up to 3568 Kbps.

[Chrysalis]

4880 vs 3568 I dont find very conclusive, they too close.

[Sephiroth]

I dug up what I'd written about rate limiting elsewhere. My take on rate limiting (I forget from where I researched this but it was in the context of corrupted downloads) was this:

The torrenting feed could exceed the rate set at the CMTS under whatever traffic shaping settings VM have set. Delay is forced to smooth the general experience out across the line card's users. You might be seeing this as speed fluctuation.

In extreme cases and if this delay piles up too much, packets are lost. So what reaches you is corrupt in so far as the checksum doesn't match what was transmitted.

O2 simply haven't set any rate limits on their routers - I would guess, or the test didn't reach the thresholds if they are set.

I would add that I don't torrent download anything so I'm relying on the Glasnost test that I haven't researched.

[jb66]

bittorrents have a tendancy to max your upload so that your download is affected

[Ignitionnet]

Fortunately not the case else the entire Internet would fall apart under congestion of any kind.

http://en.wikipedia.org/wiki/TCP_congestion_avoidance_algorithm

And given we're dealing with uTorrent which uses uTP:

http://bittorrent.org/beps/bep_0029.html

Fairly basic stuff.

[jb66]

would utp work with multiple computers sharing the same router?

[Ignitionnet]

Sure.