dslreports logo
site
 
    All Forums Hot Topics Gallery
spc

spacer




how-to block ads


Search Topic:
uniqs
1471
share rss forum feed


mb

join:2000-07-23
Washington, NJ
Reviews:
·Comcast
·Verizon Online DSL

[IPv6] IPv4 Upload slower than IPv6

Click for full size
Click for full size
I have a 50/10 Blast connection. A variety of speed tests always returned expected results until recently. Now various speed tests consistently show approximately 50/6 with the exception of the Comcast's IPv6 sight which always reports 50/10. I've attached two screenshots of tests to Comcast's Plainfield NJ IPv6 and IPv4 servers that I just conducted which show this anomoly. Any ideas why IPv6 uploads are running faster?
--
"When will they ever learn? When will they ever learn?"
Pete Seeger 1961



YukonHawk

join:2001-01-07
Patterson, NY

Maybe less congestion?



whfsdude
Premium
join:2003-04-05
Washington, DC
reply to mb

In most cases you will not see a difference in speed at such low speeds or without carrier grade NAT present.

However jitter should be less on any v6 native connection where even a single NAT instance exists.



beachintech
There's sand in my tool bag
Premium
join:2008-01-06
kudos:5
Reviews:
·Comcast
reply to YukonHawk

said by YukonHawk:

Maybe less congestion?

It's the same network, and same server.
--
Ex-Comcast Tech at the Beach. I speak for myself, not my former employer.


YukonHawk

join:2001-01-07
Patterson, NY

Makes sense.



NetFixer
From my cold dead hands
Premium
join:2004-06-24
The Boro
Reviews:
·Cingular Wireless
·Comcast Business..
·Vonage
·Comcast
reply to beachintech

said by beachintech:

said by YukonHawk:

Maybe less congestion?

It's the same network, and same server.

However, the routing can be different, and that might make a difference for some customers.

Here is the routing using the Comcast 6to4 tunnel from my current connection:


C:\>tracert speedtest.comcast.net
 
Tracing route to speedtest.g.comcast.net [68.87.73.52]
over a maximum of 30 hops:
 
  1    <1 ms     1 ms    <1 ms  ap2.dcs-net [192.168.10.1]
  2    23 ms    14 ms    10 ms  67.177.172.1
  3    14 ms    11 ms    29 ms  xe-4-0-0-0-sur01.murfreesboro.tn.nash.comcast.net [68.85.50.125]
  4    15 ms    17 ms    10 ms  162-151-9-17-static.hfc.comcastbusiness.net [162.151.9.17]
  5    12 ms    11 ms    12 ms  ae-3-0-ar01.goodslettvll.tn.nash.comcast.net [68.86.148.69]
  6    21 ms    22 ms    20 ms  68.86.94.125
  7    44 ms    34 ms    35 ms  he-0-8-0-0-cr01.ashburn.va.ibone.comcast.net [68.86.89.169]
  8    35 ms    33 ms    36 ms  so-1-1-0-0-ar04.capitolhghts.md.bad.comcast.net [68.86.90.246]
  9    37 ms    35 ms    34 ms  te-2-1-ur03.manassascc.va.bad.comcast.net [68.85.130.29]
 10    48 ms    42 ms    49 ms  speed-upload-01.manassascc.va.bad.comcast.net [68.87.73.52]
 
Trace complete.
 
C:\>tracert ipv6.speedtest.comcast.net
 
Tracing route to ipv6.speedtest.g.comcast.net [2001:558:1010:5:68:87:73:52]
over a maximum of 30 hops:
 
  1     1 ms    <1 ms    <1 ms  2002:43b1:ad12:0:a221:b7ff:fe9c:602
  2     *       23 ms    22 ms  2002:c058:6301::
  3    22 ms    22 ms    20 ms  ge-7-2-ur02.s3ndigital.ga.atlanta.comcast.net [2001:558:fe12:1::1]
  4    23 ms    23 ms    20 ms  xe-2-0-1-0-ar01.d1stonemtn.ga.atlanta.comcast.net [2001:558:140:22::1]
  5    37 ms    37 ms    38 ms  pos-2-14-0-0-cr01.350ecermak.il.ibone.comcast.net [2001:558:0:f73d::1]
  6     *        *       65 ms  he-4-4-0-0-cr01.newyork.ny.ibone.comcast.net [2001:558:0:f624::1]
  7    87 ms    56 ms    57 ms  so-1-0-0-0-ar04.whitemarsh.md.bad.comcast.net [2001:558:0:f693::2]
  8    62 ms    65 ms    63 ms  te-2-1-ur04-d.manassascc.va.bad.comcast.net [2001:558:340:c::1]
  9    62 ms    62 ms    61 ms  2001:558:1010:5:68:87:73:52
 
Trace complete.
 


Here is the routing using the Comcast native dual stack from my current connection:


C:\>tracert speedtest.comcast.net
 
Tracing route to speedtest.g.comcast.net [69.252.23.68]
over a maximum of 30 hops:
 
  1    <1 ms    <1 ms    <1 ms  ap2.dcs-net [192.168.10.1]
  2    39 ms    10 ms    13 ms  67.177.172.1
  3    14 ms     9 ms     9 ms  xe-4-0-0-0-sur01.murfreesboro.tn.nash.comcast.net [68.85.50.125]
  4    13 ms    10 ms    12 ms  xe-4-1-1-0-ar03.nashville.tn.nash.comcast.net [68.86.176.101]
  5    13 ms    11 ms    10 ms  ae-3-0-ar01.goodslettvll.tn.nash.comcast.net [68.86.148.69]
  6    24 ms    39 ms    25 ms  pos-5-6-0-0-cr01.56marietta.ga.ibone.comcast.net [68.86.90.89]
  7    35 ms    35 ms    35 ms  he-2-14-0-0-cr01.chicago.il.ibone.comcast.net [68.86.87.113]
  8    33 ms    36 ms    34 ms  he-1-6-0-0-cr01.350ecermak.il.ibone.comcast.net [68.86.88.42]
  9    45 ms    43 ms    47 ms  pos-0-13-0-0-ar01.roseville.mn.minn.comcast.net [68.86.94.74]
 10    46 ms    43 ms    42 ms  pos-0-1-0-0-ar01.crosstown.mn.minn.comcast.net [68.87.174.1]
 11    43 ms    42 ms    41 ms  te-2-3-ur02.roseville.mn.minn.comcast.net [68.87.174.190]
 12    42 ms    40 ms    40 ms  speed-upload-01.roseville.mn.minn.comcast.net [69.252.23.68]
 
Trace complete.
 
C:\>tracert ipv6.speedtest.comcast.net
 
Tracing route to ipv6.speedtest.g.comcast.net [2001:558:1010:5:68:87:73:52]
over a maximum of 30 hops:
 
  1    <1 ms    <1 ms    <1 ms  2601:5:c80:56:a221:b7ff:fe9c:602
  2    16 ms    19 ms    22 ms  2001:558:4013:19::1
  3    11 ms    10 ms    11 ms  xe-4-0-0-0-sur01.murfreesboro.tn.nash.comcast.net [2001:558:162:32::1]
  4    12 ms    11 ms    20 ms  xe-3-1-2-0-ar03.nashville.tn.nash.comcast.net [2001:558:160:a::1]
  5    26 ms    23 ms    23 ms  pos-2-10-0-0-cr01.350ecermak.il.ibone.comcast.net [2001:558:0:f7f4::1]
  6    51 ms    47 ms    47 ms  he-4-1-0-0-cr01.newyork.ny.ibone.comcast.net [2001:558:0:f621::1]
  7    47 ms    46 ms    48 ms  2001:558:0:f78f::2
  8    53 ms    52 ms    53 ms  te-2-1-ur04-d.manassascc.va.bad.comcast.net [2001:558:340:c::1]
  9    53 ms    52 ms    52 ms  2001:558:1010:5:68:87:73:52
 
Trace complete.
 



However, even though the IPv4/IPv6 paths are radically different from my connection I don't see any big difference between IPv4 and IPv6 for the Comcast speed test site. On any given hour/day, the results may vary slightly, but I don't see any pronounced tendency for faster results from either test.







--
A well-regulated militia, being necessary to the security of a free State, the right of the people to keep and bear arms shall not be infringed.

When governments fear people, there is liberty. When the people fear the government, there is tyranny.


mb

join:2000-07-23
Washington, NJ

1 edit

Please explain how to determine the URL or IP address of specific Comcast test servers so I can run some tests. Plainfield NJ seems closest to me. In the interim I have tested to the servers referenced in your tests.



mb

join:2000-07-23
Washington, NJ
Reviews:
·Comcast
·Verizon Online DSL

Click for full size
Click for full size
Below are both trace routes and pings to Comcast's IPv4 and IPv6 test servers in what appears to be Roseville MN. I have also attached images of Comcast's IPv4 and IPv6 speed test results to their servers in Twin Cities MN, which I'm guessing are the ones that the trace routes and pings are showing. Although the routing is similar, and the pings are nearly identical, IPv6 upload speeds are still clearly better than IPv4 uploads to the same servers at the same times.

C:\Users\Mark>tracert speedtest.g.comcast.net

Tracing route to speedtest.g.comcast.net [69.252.23.68]
over a maximum of 30 hops:

1 1 ms 1 ms 1 ms Cisco51493 [192.168.1.1]
2 40 ms 39 ms 20 ms c-69-141-60-1.hsd1.nj.comcast.net [69.141.60.1]
3 158 ms 67 ms 9 ms te-1-2-ur02.portmurray.nj.panjde.comcast.net [68.85.152.157]
4 10 ms 9 ms 9 ms te-9-3-ur01.portmurray.nj.panjde.comcast.net [68.86.210.213]
5 11 ms 12 ms 10 ms xe-8-1-3-0-ar03.plainfield.nj.panjde.comcast.net [68.86.209.137]
6 13 ms 13 ms 14 ms pos-3-8-0-0-cr01.newyork.ny.ibone.comcast.net [68.86.95.177]
7 39 ms 35 ms 36 ms he-0-2-0-0-cr01.350ecermak.il.ibone.comcast.net [68.86.88.138]
8 41 ms 39 ms 43 ms pos-0-14-0-0-ar01.roseville.mn.minn.comcast.net [68.86.94.78]
9 42 ms 44 ms 46 ms pos-0-1-0-0-ar01.crosstown.mn.minn.comcast.net [68.87.174.1]
10 42 ms 42 ms 43 ms te-2-1-ur02.roseville.mn.minn.comcast.net [68.87.174.18]
11 42 ms 41 ms 41 ms speed-upload-01.roseville.mn.minn.comcast.net [69.252.23.68]

Trace complete.

C:\Users\Mark>tracert ipv6.speedtest.comcast.net

Tracing route to ipv6.speedtest.g.comcast.net [2001:558:1030:8:69:252:23:68]
over a maximum of 30 hops:

1 1 ms 1 ms 1 ms 2601:c:3180:62:5a6d:8fff:----:----
2 21 ms 28 ms 20 ms 2001:558:6026:67::1
3 10 ms 9 ms 9 ms te-1-2-ur01.portmurray.nj.panjde.comcast.net [2001:558:12:64::1]
4 12 ms 10 ms 11 ms xe-8-1-3-0-ar03.plainfield.nj.panjde.comcast.net [2001:558:10:233::1]
5 12 ms 17 ms 14 ms pos-3-10-0-0-cr01.newyork.ny.ibone.comcast.net [2001:558:0:f736::1]
6 43 ms 35 ms 35 ms he-0-0-0-0-cr01.350ecermak.il.ibone.comcast.net [2001:558:0:f620::2]
7 44 ms 43 ms 43 ms pos-0-15-0-0-ar01.roseville.mn.minn.comcast.net [2001:558:0:f794::2]
8 61 ms 59 ms 59 ms te-0-3-0-11-ar01.crosstown.mn.minn.comcast.net [2001:558:30:d1::1]
9 60 ms 58 ms 57 ms te-2-3-ur02.roseville.mn.minn.comcast.net [2001:558:30:74::2]
10 57 ms 58 ms 57 ms 2001:558:1030:8:69:252:23:68

Trace complete.

Pinging speedtest.g.comcast.net [68.87.73.52] with 32 bytes of data:
Reply from 68.87.73.52: bytes=32 time=20ms TTL=57
Reply from 68.87.73.52: bytes=32 time=21ms TTL=57
Reply from 68.87.73.52: bytes=32 time=20ms TTL=57
Reply from 68.87.73.52: bytes=32 time=21ms TTL=57

Ping statistics for 68.87.73.52:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 20ms, Maximum = 21ms, Average = 20ms

C:\Users\Mark>ping ipv6.speedtest.g.comcast.net

Pinging ipv6.speedtest.g.comcast.net [2001:558:1010:5:68:87:73:52] with 32 bytes
of data:
Reply from 2001:558:1010:5:68:87:73:52: time=21ms
Reply from 2001:558:1010:5:68:87:73:52: time=20ms
Reply from 2001:558:1010:5:68:87:73:52: time=20ms
Reply from 2001:558:1010:5:68:87:73:52: time=19ms

Ping statistics for 2001:558:1010:5:68:87:73:52:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 19ms, Maximum = 21ms, Average = 20ms
--
"When will they ever learn? When will they ever learn?"
Pete Seeger 1961



NetFixer
From my cold dead hands
Premium
join:2004-06-24
The Boro
Reviews:
·Cingular Wireless
·Comcast Business..
·Vonage
·Comcast
reply to mb

said by mb:

Please explain how to determine the URL or IP address of specific Comcast test servers and I will post the trace routes. Plainfield NJ seems closest to me.

I did not bother with that with the results I posted (I was only illustrating that the IPv4 and IPv6 paths to the same server were not necessarily even close), but you can monitor the browser status bar to see what current URL(s) are being used. Shown below are a couple of examples I just did:







The URLs I saw when using the Plainfield server were (the results may change depending on your own IP address, and the server load):

http://plainfield.speedtest.comcast.net
http://plainfield2.speedtest.comcast.net
http://ipv6.plainfield.speedtest.comcast.net
http://ipv6.plainfield2.speedtest.comcast.net
--
A well-regulated militia, being necessary to the security of a free State, the right of the people to keep and bear arms shall not be infringed.

When governments fear people, there is liberty. When the people fear the government, there is tyranny.


mb

join:2000-07-23
Washington, NJ

1 edit

Anything standout in the trace routes to the two Minnesota servers that I posted above? If anything, although there is an additional hop in the IPv4 route, the IPv6 route looks to have more latency.



NetFixer
From my cold dead hands
Premium
join:2004-06-24
The Boro
Reviews:
·Cingular Wireless
·Comcast Business..
·Vonage
·Comcast

said by mb:

Anything standout in the trace routes to the two Minnesota servers that I posted above, if anything, although there is an additional hop in the IPv4 route, the IPv6 routes looks to have more latency.

Nothing stands out to me. The IPv4 and IPv6 traceroutes I did took very different geographic paths to the endpoint, but yours seem to be similar.

You can try the same comparison test to the plainfield servers to see if there is any difference there. The thing to look for would likely be an obviously different geographic path, not necessarily different lag times (because ICMP echo packets have different priority than the TCP payload packets used by the test, and you would be making an apples vs oranges comparison).

The bottom line is that you may never know why the IPv6 server is returning a faster result than the IPv4 server. Is it just that server that produces that symptom, or do all/most of the Comcast speed test servers do the same for your connection?
--
A well-regulated militia, being necessary to the security of a free State, the right of the people to keep and bear arms shall not be infringed.

When governments fear people, there is liberty. When the people fear the government, there is tyranny.


mb

join:2000-07-23
Washington, NJ
Reviews:
·Comcast
·Verizon Online DSL

1 edit

All (continental US) IPv4 test servers now come in between 5 and 7 up and ShaperProbe is showing the same... for some reason IPv4 upload speeds have decreased but IPv6 upload speed have not. Regardless of which Comcast servers I am using, I am always seeing about a 3 to 4 improvement on the IPv6 upload speed when using their same sites. Anyone know of any non-Comcast IPv6 speedtests?

I've got a couple of IM's in to Comcast insiders but haven't heard anything back as of yet.
--
"When will they ever learn? When will they ever learn?"
Pete Seeger 1961



NetFixer
From my cold dead hands
Premium
join:2004-06-24
The Boro
Reviews:
·Cingular Wireless
·Comcast Business..
·Vonage
·Comcast

said by mb:

All (continental US) IPv4 test servers now come in between 5 and 7 up and ShaperProbe is showing the same... for some reason IPv4 upload speeds have decreased but IPv6 upload speed have not. Regardless of which Comcast servers I am using, I am always seeing about a 3 to 4 improvement on the IPv6 upload speed when using their same sites. Anyone know of any non-Comcast IPv6 speedtests?

I've got a couple of IM's in to Comcast insiders but haven't heard anything back as of yet.

I'm sure that eventually the IPv6 results will change to match the IPv4 results.



--
A well-regulated militia, being necessary to the security of a free State, the right of the people to keep and bear arms shall not be infringed.

When governments fear people, there is liberty. When the people fear the government, there is tyranny.


jlivingood
Premium,VIP
join:2007-10-28
Philadelphia, PA
kudos:2
reply to mb

I would hope that IPv6 will be a bit better. But it is important to bear in mind that v6 support on our Speed Test site is still in beta - which you'll notice for things like showing your IPv6 address but the result saying IPv4.
--
JL
Comcast



mb

join:2000-07-23
Washington, NJ
Reviews:
·Comcast
·Verizon Online DSL
reply to mb

Re: [IPv6] IPv4 Upload slower than IPv6 *Issue Resolved*

This issue has been resolved. Thanks to all for your interest and input. The issue appears to have been related to the network traffic examination feature provided by Avast security software which had been recently installed. Upon it's removal and the reinstallation of Avira protection software things are back to normal.