Is IPv6 faster than IPv4?

[Hellfire323]

I currently have a Cisco DPC3008 modem with a Linksys E2500 Wireless router with Comcast Xfinity internet rated at 50mb/s. The computer directly attached to the wireless router via ethernet consistently receives over 50mb/s using various internet speed testing sites, including Comcast's own speed tester.

My other computer is located in another room and is approximately 15 feet away from the router and behind one wall. I recently bought a ASUS PCE-AC56 wireless adapter that is capable of 802.11ac. I disabled 802.11b/g on the wireless adapter and set my router to 802.11 N only over the 5 ghz channel.

Now my computer receives about 24mb/s over the wireless LAN, but I cannot get it to achieve the full 50mb/s. There are no devices between my computer and wireless router. I used comcast's own speed testing site, the results it gave me are:
IPv4 - 26.7mb/s download & 11.3mbs upload
IPv6 - 56.6mb/s download & 11..5mb/s upload

I was under the impression that the IPv6 was just an IP address scheme to replace the limited amount of IPv4 addresses, but does it also affect internet speed? And is there a way to increase the IPv4 speed or does it even matter?

 

[ronintexas]

IP Addresses are only a method by which you can identify other computers on the "network" (the internet is a network). You will find that at different times of the day (even seconds apart), you will get different results on your tests.

Theoretically, you should get a steady 50Mbps down and 5 Mbps up with your setup. I will say that with wireless connections - you almost never get the full upload/download rate. My laptop sits on my desk (router it 10 feet away), and I never get full speed unless I am connected via ethernet.

Austin has crazy speed.....this is TWC: http://www.speedtest.net/my-result/3698436112

I get anywhere between 280Mbps to 330Mbps down and 18Mbps - 24Mbps up depending upon the time of day.

In some circumstances, there may be interference with neighbor's equipment, you might try a different channel for the router. Anything electronic in the house can interfere also - microwaves, other devices, ceiling fans, appliances, vacuum cleaners, etc.....

 

[Hellfire323]

IP Addresses are only a method by which you can identify other computers on the "network" (the internet is a network). You will find that at different times of the day (even seconds apart), you will get different results on your tests.

Theoretically, you should get a steady 50Mbps down and 5 Mbps up with your setup. I will say that with wireless connections - you almost never get the full upload/download rate. My laptop sits on my desk (router it 10 feet away), and I never get full speed unless I am connected via ethernet.

Austin has crazy speed.....this is TWC: http://www.speedtest.net/my-result/3698436112

I get anywhere between 280Mbps to 330Mbps down and 18Mbps - 24Mbps up depending upon the time of day.

In some circumstances, there may be interference with neighbor's equipment, you might try a different channel for the router. Anything electronic in the house can interfere also - microwaves, other devices, ceiling fans, appliances, vacuum cleaners, etc.....

So in theory, if I unplugged every electronic device in the house and moved my computer closer to the router, I might get faster speeds? Everything I read always mentions interference as the main cause of speed degradation on the wifi, but my router is the only device on channel 149 on the 5ghz range. On paper it would seem my wireless connected computer should get 50mb/s (and it does on the IPv6 test according to Comcast). Perhaps this is an ISP issue? Would the ISP purposely throttle down IPv4 traffic?

 

[bill001g]

The big question would be are they actually the same site or different sites that they use for ipv4 and ipv6. If they are not the exact same server in the exact same data center you can not compare the numbers. I would bet they are not the same machine.

Now lets assume they are the same. A IPv6 connection should actually get LESS throughput than IPv4 but it depends how you measure. There are a minimum of 20 extra bytes of header in a IPv6 packet and the MTU has not been increased on most networks. This means you must now send more packets with more headers to transfer the same amount of data. So if you look at the actual usable data a IPv6 must transfer less data than a IPv4 on the same exact connection because there is more overhead. Of course if you include all the overhead then they should be exactly the same but how useful is that.

 

[Hellfire323]

The big question would be are they actually the same site or different sites that they use for ipv4 and ipv6. If they are not the exact same server in the exact same data center you can not compare the numbers. I would bet they are not the same machine.
.

To my knowledge the server used in Comast's speed test is the same server, or at least within the same location, considering you select the site of the server before the test begins. With that said, it may not be the same exact physical device.

http://stage.results.speedtest.comcast.net/result/601251344.png

Shows my latest test results, 26.07mb/s download over IPv4 but 56.39mb/s over IPv6. I still haven't been able to figure this out. I'm all out of ideas.

 

[bigpinkdragon286]

Perhaps the server handling the IPv6 speed test has far fewer people making requests of it than the IPv4 server?

Perhaps your router handles the traffic differently?

 

[Hellfire323]

So for anyone keeping up with this post, I finally figured everything out. First of all, I challenge everyone to always fully investigate issues before succumbing to the "most probable" problem. Sure interference plays a huge part in speed degradation over the wifi, but with current technology your wireless speed should be on par with your wired. (considering you're taking full advantage of the newer technology.)

The problem begins with Comcast, specifically, the QoS mechanisms on their traffic. http://forums.comcast.com/t5/Home-Networking-Router-WiFi/IPv4-Traffic-Slow-Native-IPv6-Full-Speed/td-p/1822405 Can explain it better than I can, but the main point is:

"*All* inbound Comcast traffic (Internet --> client) was tagged with a DSCP value of 8 (Class Selector 1). The DSL traffic had a DSCP value of 0. So Comcast is tagging all traffic to be treated a certain way by QoS: "Priority," which sounds good but is actually the second-*lowest* possible."

Furthermore, the WMM on my Linksys E series router (E2500) treated the IPv4 Comcast traffic poorly and I believe was causing a slower speed over the IPv4 traffic. In order to troubleshoot I decided to buy a new router (i tried the DD-WRT firmware, but it didn't support the 5Ghz channel on my make/model of router.) and bought an ASUS RT-AC66U. I now get the advertised speed on both my wired and wireless computers. http://stage.results.speedtest.comcast.net/result/607461424.png 57.7mb/s - download over wifi.

If you're having a similar problem, I suggest using a packet sniffer such as Wire Shark to look for the DSCP value of the incoming packets.

Thanks everyone for the input and hopefully this helps someone out.

 

[bigpinkdragon286]

While I don't claim to have a definitive answer, I'm skeptical that Comcast's setting of a particular priority level is the source of the problem. Both wired and wireless should be subject to the same QoS, if they're going through the same router.

It sounds to me like, you bought a different wireless router and that solved your problem? Leads me to suspect the original piece of equipment was the source of the problem, whether that be bad programming, bad or inadequate hardware for the task, or just improper configuration.

Looking at QoS priority levels doesn't sound like a solution to me since properly working QoS on both routers should prioritize your packets the same way. Why not test your wireless router with QoS turned off then, if you suspect it as the problem?

On the other hand, if you are only running a single stream, to test your download speeds, what prioritizing of packets is your QoS feature even doing that might decrease your speed? Correct me if I'm wrong but, unless you're running another task on the network for QoS to be prioritizing above your speed test, or you have QoS implementing some sort of speed limiting or throttling, what is QoS actually doing at this point to affect your traffic?

I'm glad you got things sorted, however, and thanks for updating so others can learn as well.

 

[Hellfire323]

....Both wired and wireless should be subject to the same QoS, if they're going through the same router.

It sounds to me like, you bought a different wireless router and that solved your problem? Leads me to suspect the original piece of equipment was the source of the problem, whether that be bad programming, bad or inadequate hardware for the task, or just improper configuration.

Looking at QoS priority levels doesn't sound like a solution to me since properly working QoS on both routers should prioritize your packets the same way. Why not test your wireless router with QoS turned off then, if you suspect it as the problem?

On the other hand, if you are only running a single stream, to test your download speeds, what prioritizing of packets is your QoS feature even doing that might decrease your speed? Correct me if I'm wrong but, unless you're running another task on the network for QoS to be prioritizing above your speed test, or you have QoS implementing some sort of speed limiting or throttling, what is QoS actually doing at this point to affect your traffic?

I'm glad you got things sorted, however, and thanks for updating so others can learn as well.

Its true the main issue was the router. "... the WMM on my Linksys E series router (E2500) treated the IPv4 Comcast traffic poorly and I believe was causing a slower speed over the IPv4 traffic." As an IT networking student, I am still studying the way QoS could have caused the issue. However, during my research I ran into several articles including this one: http://www.dd-wrt.com/wiki/index.php/Comcast_download_speed_fix_for_Linksys_eSeries that explains that the Linksys E-series router (which I had) had an issue with Comcast traffic and its DSCP setting.

I was limited in the fact that I did not have another ISP line to test, however, in one of the articles, someone did have a DSL line and compared the Comcast with the AT&T DSL line. The link for that article is in one of the previous post. I could not turn the WMM QoS feature off, as you need it to run 802.11n. I needed the -n- protocol in order to achieve a fast enough speed in order to more accurately test the internet speed. The use of a different firmware such as the DD-WRT was also suggested, but as I stated, it does not support the faster 5ghz channels.

But yes the router was the issue. More specifically, it was how the router was handling the Comcast traffic, seemingly due to the DSCP setting set on the Comcast traffic. Whether or not this is 100% accurate, I do not know. But I do know it, after hours of research, this is the most probable cause of the problem and after buying a new router, the issue was fixed.

 

[bill001g]

Although the DSCP story sound plausible it likely was never your problem. QoS never drops traffic just because of DSCP setting. It will only drop traffic if it must make a choice between 2 packets and its buffers are full that it uses to temporarily store excess traffic. A wireless network is many times faster than almost any internet connection so it would be practically impossible to receive packets from a internet connection faster than they could be sent out on wireless. Since there will not be any contention or queuing of packets the router should not be put in a position to have to select which packet to drop.

Pretty much you would only use the WMM feature if you had say lan to wireless traffic that could send burst of traffic very quickly.

Hard to say it could have been another bug in the WMM for that vendor. WMM is mostly ineffective even when it is working. Sure the router may be able to control the order it sends data but it has no control over any other device. Interference both from devices associated with the router and random other radios cause massive data loss and delay that no QoS can solve.

 

[christinebcw]

Bill, thanks also for the header-size and packet-size reminder.

 

[ronintexas]

I don't have the fastest ISP in Austin (currently ranked #4 in speed).....

http://www.speedtest.net/my-result/3716696596

 

[Hellfire323]

Although the DSCP story sound plausible it likely was never your problem. QoS never drops traffic just because of DSCP setting. It will only drop traffic if it must make a choice between 2 packets and its buffers are full that it uses to temporarily store excess traffic. A wireless network is many times faster than almost any internet connection so it would be practically impossible to receive packets from a internet connection faster than they could be sent out on wireless. Since there will not be any contention or queuing of packets the router should not be put in a position to have to select which packet to drop.

Pretty much you would only use the WMM feature if you had say lan to wireless traffic that could send burst of traffic very quickly.

Hard to say it could have been another bug in the WMM for that vendor. WMM is mostly ineffective even when it is working. Sure the router may be able to control the order it sends data but it has no control over any other device. Interference both from devices associated with the router and random other radios cause massive data loss and delay that no QoS can solve.

It's very possible that the DSCP setting had nothing to do with my issue. But remember, in the original post I was receiving my full advertised speed over the IPv6 traffic even with the higher number of packets because of the 1500 bytes MTU cap. Assuming that the test server was the exact same device sending both IPv4 and IPv6 traffic; this meant some device was treating the IPv4 traffic differently. When I came across this article http://forums.comcast.com/t5/Home-Networking-Router-WiFi/IPv4-Traffic-Slow-Native-IPv6-Full-Speed/td-p/1822405 I found that Mr. Hilliard was having a very similar issue with the same E-series Linksys router.

What Mr. Hilliard found:
"*All* inbound Comcast traffic (Internet --> client) was tagged with a DSCP value of 8 (Class Selector 1). The DSL traffic had a DSCP value of 0. So Comcast is tagging all traffic to be treated a certain way by QoS: "Priority," which sounds good but is actually the second-*lowest* possible. "

"WMM, itself a QoS technique, apparently de-prioritizes (drops?) based on the Comcast-supplied value. Turning off WMM worked around it - but since WMM is part of the 802.11n spec, I wanted root cause. Judiciously replacing that set-by-Comcast DSCP value does the trick. "

"So between my Linksys router and both ISPs, I had a Netscreen firewall. It lets me set DSCP values by policy - so I told it to match the DSL (DSCP 0). This yielded great improvement. However, I was still not getting full speed so even a zero value was not the best for > DSL rates. I set the DSCP value to 46 (Expedited Forwarding) and bingo, up to 20Mbps, almost full provisioned speed (25Mbps). "

"IPv6 worked for me was because DSCP was configured correctly for that but it is incorrect for IPv4. Since the only packets with this bad information are the ones that are being downloaded upload is largely unaffected.""

The very last post on that page is a user with another similar problem and again is using a Linksys E2100 router. He said he was able to solve his issue by switching to 802.11g, which does not need the WMM to work. MY conclusion was that my Linksys E-series router, for whatever reason, was not giving me the speed I was paying for, so it has since been replaced. Also, WMM may decide to delay network traffic, which it may have been doing with the IPv4 traffic. I will continue to investigate the issue.

 

[Hellfire323]

I've used WireShark to capture both IPv4 and IPv6 traffic from http://speedtest.comcast.net/

Unfortunately, I can't quite make heads or tails of the packet capture, considering I'm not sure what is the lowest/highest priority. From what I have found, the class selector 1 (0x08) is the lowest possible priority, while IPv6's 0 means that it is treated as normal traffic. Mr. Hilliard (also known as Lightsword on the web site below) is adamant in suggesting the (0x08) is the lowest priority setting, which is why Linksys routers faulty WMM mechanism is delaying traffic of the IPv4. I would really appreciate if someone could explain to me what exactly all these numbers mean on the captured packets.

http://www.linksysinfo.org/index.php?threads/critical-dscp-bug-affecting-wifi-download-speeds-on-comcast-cisco-e-series-and-others.69006/

 

[bill001g]

The actual function of the marking varies by the chipset manufacture. But let say priority 1 is low and priority 0 is normal which is what the general tables say.

So if I were sending lots and lots or priority 0 traffic and a priority 1 packet needed be sent it would drop that packet. So in this case it traffic with 0 would be preferred over traffic with a 1.

Now lets say I am only sending priority 1 traffic like your test should have been...assuming you were not simultaneously testing ipv6 and ipv4. Why would it drop any of it. Its not like it needs to send a priority 0 packet.

Although you can configure hard limits on DSCP on a traffic shaper or commercial router there is no cap on some certain DSCP value. It does not say if they traffic is marked xx it can only be 5m/sec.

So it makes no sense that if all you were doing is running just the speed test and there was no other traffic it would cause any form of limitation no matter what the packets are marked.

All you can tell for sure is some manufacture had some bug in their software. If the bug causes the traffic to be limited when there is no contention for bandwidth then you could see a symptom like this but WMM is not suppose to work that way. Like most QoS it will not limit traffic if there is no other competing traffic.

 

[ronintexas]

Funny - IPV4 was faster for me: http://stage.results.speedtest.comcast.net/result/610189004.png

 

[orionite]

So for anyone keeping up with this post, I finally figured everything out. First of all, I challenge everyone to always fully investigate issues before succumbing to the "most probable" problem. Sure interference plays a huge part in speed degradation over the wifi, but with current technology your wireless speed should be on par with your wired. (considering you're taking full advantage of the newer technology.)

The problem begins with Comcast, specifically, the QoS mechanisms on their traffic. http://forums.comcast.com/t5/Home-Networking-Router-WiFi/IPv4-Traffic-Slow-Native-IPv6-Full-Speed/td-p/1822405 Can explain it better than I can, but the main point is:

"*All* inbound Comcast traffic (Internet --> client) was tagged with a DSCP value of 8 (Class Selector 1). The DSL traffic had a DSCP value of 0. So Comcast is tagging all traffic to be treated a certain way by QoS: "Priority," which sounds good but is actually the second-*lowest* possible."

Furthermore, the WMM on my Linksys E series router (E2500) treated the IPv4 Comcast traffic poorly and I believe was causing a slower speed over the IPv4 traffic. In order to troubleshoot I decided to buy a new router (i tried the DD-WRT firmware, but it didn't support the 5Ghz channel on my make/model of router.) and bought an ASUS RT-AC66U. I now get the advertised speed on both my wired and wireless computers. http://stage.results.speedtest.comcast.net/result/607461424.png 57.7mb/s - download over wifi.

If you're having a similar problem, I suggest using a packet sniffer such as Wire Shark to look for the DSCP value of the incoming packets.

Thanks everyone for the input and hopefully this helps someone out.

Hi, I have an extreme case of ipv4 vs ipv 6 discrepancy. http://results.speedtest.comcast.net/result/909781664.png

I also, have the exact same router, you have, but it doesn't seem to matter where I plug in my cable, router or modem directly, I'm usually getting insanely low ipv4 speeds and great (as advertised) ipv6 speeds.

Comcast said WMM was actuall switched off on my modem. The tech that came to my house checked cables outside the house. Then he plugged in his laptop into the modem and got great ipv6 speeds. I plugged in my pc, got crap. So I bought a new cable (cat6), hoping it would make a difference. Doesn't seem to.

Can you share any settings you made in your Asus Router that might help? I'm really at a loss here, and don't know what else to try other than maybe buying a dedicated network card. Btw, I downloaded wireshark, but frankly, I don't even know where to begin there. Those acronyms and numbers mean nothing to me.