dslreports logo
 
    All Forums Hot Topics Gallery
spc
Search similar:


uniqs
6890

cp
Premium Member
join:2004-05-14
Wheaton, IL

cp

Premium Member

[Speed] IPv4 vs IPv6 Speed Test

I'm curious how people's speed tests are comparing against the two protocols. From my observations, IPv6 always seems to be a tad slower than IPv4.

IPv4:



IPv6:



Also, Comcast needs to change their v6 speed test so the label actually reads IPv6.

Streetlight
join:2005-11-07
Colorado Springs, CO

1 edit

Streetlight

Member

Deleted my post - misread what you wanted.

Sorry
Dan2112
join:2001-08-24
San Jose, CA

Dan2112 to cp

Member

to cp
Similar results, but not by much:

IPv4:



Ipv6:


whfsdude
Premium Member
join:2003-04-05
Washington, DC

whfsdude to cp

Premium Member

to cp
I'm using my connection a fair bit atm so these aren't capturing my full speeds.

IPv4:


IPv6:
nysports4evr
Premium Member
join:2010-01-23

1 recommendation

nysports4evr

Premium Member

said by whfsdude:

I'm using my connection a fair bit atm so these aren't capturing my full speeds.

Showoff.
Emiya
join:2006-03-30
Southington, OH

Emiya to cp

Member

to cp
Curious, there seems to be a discrepancy on the server distance for the same cities.

whfsdude
Premium Member
join:2003-04-05
Washington, DC

whfsdude

Premium Member

said by Emiya:

Curious, there seems to be a discrepancy on the server distance for the same cities.

IP geolocation information. Many of the v6 allocations are newer which is likely why.
whfsdude

whfsdude to nysports4evr

Premium Member

to nysports4evr
said by nysports4evr:

said by whfsdude:

I'm using my connection a fair bit atm so these aren't capturing my full speeds.

Showoff.

Ha! I'm probably the happiest Comcast customer there is.

Paused most of my traffic so this should be more accurate.

IPv4:


IPv6:

cp
Premium Member
join:2004-05-14
Wheaton, IL

cp to Emiya

Premium Member

to Emiya
My geolocation information is definitely off for my v4.

I just purchased a new router this week and Comcast gave me a new v4 ip because of that. The geolocation for my v4 shows up in Kansas somewhere instead of Chicago.

I think v6 is going to be off for a long time for everyone.
cp

cp to whfsdude

Premium Member

to whfsdude
Yeah, I would be too lol!

Sadly Comcast won't start offering that service in my area until there's a viable competitor that would offer those speeds. No FIOS here. Just U-Verse and Comcast.

Even if they did, I doubt I'd fork over the $$$ required for it.

Tobin
join:2003-09-21
Burlingame, CA

1 edit

Tobin to cp

Member

to cp
I'm getting lower IPv6 results as well, but bear in mind there are so many possible reasons for this. For example, my results differ just based on which Flash implementation I use, (Windows 8 Explorer, Chrome or Firefox style plugins...) On Chrome and Explorer, my ping is 9ms instead of 11-12ms on Firefox. On Chrome my upload via IPv6 is barely 9mbps whereas every other result is >11mbps.

IPv6 slowdowns could be the result of Ooka's own software, hardware at each individual site, etc...

whfsdude
Premium Member
join:2003-04-05
Washington, DC

whfsdude to cp

Premium Member

to cp
Kind of off-topic but "Washington DC, VA" LOL! It's either Ashburn, VA or Washington, DC.

Darknessfall
Premium Member
join:2012-08-17
Motorola MG8725
Asus RT-N66

Darknessfall to Tobin

Premium Member

to Tobin
said by Tobin:

I'm getting lower IPv6 results as well, but bear in mind there are so many possible reasons for this. For example, my results differ just based on which Flash implementation I use, (Windows 8 Explorer, Chrome or Firefox style plugins...) On Chrome and Explorer, my ping is 9ms instead of 11-12ms on Firefox. On Chrome my upload via IPv6 is barely 9mbps whereas every other result is >11mbps.

IPv6 slowdowns could be the result of Ooka's own software, hardware at each individual site, etc...

IE and Chrome usually have lower and incorrect pings at least in my experience. I'm sure I don't have the 5ms latency to CT to Boston, Mass as it claims lol. Firefox and tracerts show about 15ms or so.

DualStack
@comcast.net

DualStack to cp

Anon

to cp
IPv4:



IPv6:


FFH5
Premium Member
join:2002-03-03
Tavistock NJ

FFH5 to cp

Premium Member

to cp
How can you be sure whether v4 or v6 is being used for the test? Different urls?

whfsdude
Premium Member
join:2003-04-05
Washington, DC

1 recommendation

whfsdude

Premium Member

said by FFH5:

How can you be sure whether v4 or v6 is being used for the test? Different urls?

Correct: »ipv6.speedtest.comcast.net

There is also this wonderful gem - for seeing when you're hitting a page over v6. It shows all the elements on the page that are v6 or v4 and if active traffic is being passed.

»chrome.google.com/websto ··· al?hl=en

FFH5
Premium Member
join:2002-03-03
Tavistock NJ

FFH5

Premium Member

V4



V6


NetFixer
From My Cold Dead Hands
Premium Member
join:2004-06-24
The Boro
Netgear CM500
Pace 5268AC
TRENDnet TEW-829DRU

NetFixer to cp

Premium Member

to cp
It depends on the time of day and the phase of the moon, but at this moment the results look like this:

IPv4:



IPv6:



This is using a business class 16/3 connection.

darklobe
@comcast.net

darklobe

Anon

said by NetFixer:

It depends on the time of day and the phase of the moon, but at this moment the results look like this:

IPv4:
[att=1]

IPv6:
[att=2]

This is using a business class 16/3 connection.

How did you get those fast upload speeds. Is that because of business class connection?

NetFixer
From My Cold Dead Hands
Premium Member
join:2004-06-24
The Boro
Netgear CM500
Pace 5268AC
TRENDnet TEW-829DRU

2 edits

NetFixer

Premium Member

said by darklobe :

How did you get those fast upload speeds. Is that because of business class connection?

It is just PowerBoost on steroids, and possibly having business class contributes to that, but based on these threads:

»[Speed] [BUSINESS CLASS] 12/2 or 22 Tier After Upgrade

»[Business] [POLL] Do you have the new 15/3 or 27/7 business spee

some business class customers now seem to have lost upstream PowerBoost completely.

It appears from my own testing to be a combination of PowerBoost and the upstream bonding algorithms that Comcast is using in the CMTS to which I connect. If I temporarily use my old DOCSIS2 D-Link DCM202 (which of course, does not have channel bonding), my upstream speed stays between 3-5 mbps, but the downstream speed still shows PowerBoost peaks of 20+ mbps. And FWIW, the old DCM202 does get a business class "c05" config file, so just having a business class connection is not the primary thing that triggers the crazy upstream PowerBoost.

My own personal preference would be (especially for business class accounts) to allow a customer selected option to either allow or disable PowerBoost. PowerBoost really complicates the job of any device or application that tries to do QoS rate limiting based on detected network speed. I understand why Comcast marketing loves PowerBoost (it is a clever marketing gimmick), but for many network applications it is a big problem, not a "free" benefit. If my old DCM202 supported native dual stack IPv4/IPv6, I would probably use it as my primary modem instead of a hot swap spare since it seems to have a more sensible PowerBoost implementation.