  kilometers
join:2007-01-24 Sacramento, CA
·SureWest Internet
·AT&T U-Verse
edit: March 14th, @02:17PM
| These bad ping times are driving me insane.
I've been dealing with it since I first had the service hoping it would somehow be fixed overtime. I understand that being on the interleaved path adds about 20ms extra to my ping time. Does anyone know if AT&T will have a solution for this anytime soon? In a lot of my online games recently my friends have been telling me I have a really bad pings (red) which is causing me to lag games. Outside gaming the connection is fine. I get my 10/1.5 speeds all the time and my TV service never goes out. But seriously, I miss comcast for its fairly low ping and non-laggy online gameplay.
I was playing Super Smash Bros. Online yesterday and I actually lagged out of two fights twice. And they said it was just me because as soon as I was gone the games continued on lag free. CoD4 I would experience major lag and die all the time because of it. All my consoles are set to DMZ and are directly conencted to the RG (except the wii uses wifi) and my connection is not being used at all when I game (TV or internet). I'd really hate to cancel my u-verse service because the TV and internet speeds are much better than what comcast gave me, but I'm a hardcore gamer and the bad latency is driving me away.
When AT&T switches to VDSL2 will it help solve the latency issue? IS there any way they could put me on the fast path just so I can see how much it effects the TV? Did at&t actuallly acknowledge there's terrible ping times when gaming on their "lightspeed" network? Does anyone else have this many problems when gaming on their U-Verse connection? I'm wondering if maybe something could just be wrong on my end and I should submit a support ticket to AT&T. |
|
  gdm Premium,MVM join:2001-06-15 Mchenry, IL clubs: | This is a known thing with u-verse. Interleaving adds latency and until AT&T feels it's ok to remove it all u-verse users are stuck on interleave. |
|
  xboxcpu
@rr.com
| reply to kilometers said by kilometers :When AT&T switches to VDSL2 will it help solve the latency issue? Who told you uverse is switching to VDSL2? |
|
  kilometers
join:2007-01-24 Sacramento, CA edit: March 14th, @04:15PM
| Actually I meant to say pair bonding. Sorry about that. |
|
 cwh
join:2006-05-14 San Antonio, TX
| reply to kilometers How bad are your pings? The interleave is only adding about 10ms to your ping times. Fastpath DSL is about 10ms on the first hop, uverse is doing about 20ms to the first hop.
You might want to do a traceroute to find where your packets are being held up. |
|
 jrb531
join:2001-02-14 Chicago, IL clubs:
edit: March 15th, @02:07PM
| First it's 20ms now it's only 10ms????
If all interleave did was add a tiny 10ms then I would have them turn it on for me and get faster service 
IMHO it's more like 20-40ms.
Do you really think that if Interleave "only" added 10ms to the latency that there would be any reason to turn it to fast?
-JB
I found this link that may help:
»www.kitz.co.uk/adsl/interleaving.htm
quote: it can increase latency by an additional 20-40ms. Whilst this will not be noticeable to the vast majority of users, ardent gamers are the ones most likely to complain about additional latency and therefore prefer a slower sync speed than higher latency.
|
|
 MyDogHsFleas Premium join:2007-08-15 Austin, TX | reply to kilometers I think your choices are:
a) dump U-verse and go to DSL (assuming it's available) or cable -- whichever'll give you the lowest latencies
b) give up Internet online gaming |
|
 cwh
join:2006-05-14 San Antonio, TX
| reply to jrb531 said by jrb531 :First it's 20ms now it's only 10ms???? If all interleave did was add a tiny 10ms then I would have them turn it on for me and get faster service  IMHO it's more like 20-40ms. Do you really think that if Interleave "only" added 10ms to the latency that there would be any reason to turn it to fast? -JB I found this link that may help: » www.kitz.co.uk/adsl/interleaving.htm quote: it can increase latency by an additional 20-40ms. Whilst this will not be noticeable to the vast majority of users, ardent gamers are the ones most likely to complain about additional latency and therefore prefer a slower sync speed than higher latency.
I can only tell you what i see. My first hop has been consistently 20ms since i got the service. This results seems fairly consistent with what others are seeing.
Normal dsl on fastpath tends to have a 10ms latency, so it appears u-verse is only getting 10ms penalty for being on interleaved. |
|
  gdm Premium,MVM join:2001-06-15 Mchenry, IL clubs:
·AT&T U-Verse
·AT&T CallVantage
·Comcast Digital Vo..
·Comcast
·Vonage
| Here is what my U-verse shows to dslreports.com
U-verse
This was from my comcast line
|
|
  kilometers
join:2007-01-24 Sacramento, CA
·SureWest Internet
·AT&T U-Verse
| reply to kilometers I tried doing a tracert to dslreports as well and these were my results. |
|
  ripperm Premium join:2004-07-17 Milwaukee, WI
·AT&T U-Verse
·AT&T Yahoo
| reply to kilometers Ping times can also be something of an issue for those of us who use remote access for work. With Elite DSL (6 mbps/768 kbps), I was able to sustain a high-res, high-color remote desktop through fairly intensive use. Although I still can with Elite on U-Verse (6 mbps/1 mbps), the additional lag is evident. On traceroutes, my first hop was 10 ms on DSL and is 21 on U-Verse.
I have also found that while speed tests like AT&Ts show me at 5.9 mbps/976 kbps, other speed test sites, perhaps relying on shorter bursts, register between 3 and 4.5 mb down. With Milwaukee Road Runner's test using a 3 mb transfer, for example, I would typically get 4.9 mbps down on my old DSL connection. On U-Verse, I get 3.25 mbps down.
We make a fairly big to-do about "high-speed" Internet, but speed is a lot more than the time it takes to download a large file (something I don't do that often), and bandwidth measures capacity more than it does speed. Latency, however, is only about speed, and when it comes to latency, U-Verse is clearly not the king of the hill.
Citing bandwidth for Internet performance is like citing CPU clock speed for computer performance. There's more to it than that. What we really need, IMHO, is a new benchmark (a "SpecMark" for Internet) that measures effective throughput across a broad and representative mix of internet-based transactions and, in the interests of "truth in advertising," for ISPs to use that instead of the increasingly meaningless bandwidth metric in their advertising. |
|
 Learux
join:2001-10-23 Valencia, CA
·AT&T U-Verse
| I took the plunge too, switched from ATT dsl to uverse DSL.
My observation is that while up/download speeds with comparable packages are a bit faster, latency deteriorates.
Test show me anywhere between 35-50ms.
Internet seems to be a bit more snappy then old style dsl, this might have been caused by my old modem/router combination
In general uverse internet is good, was going to hang up my gaming hat anyway.
The only type of gaming it really affects is First person shooters.
Learux |
|
  apeface
join:2000-09-16 Mckinney, TX | reply to kilometers I've been playing COD, COH, BF2142, and WIC on both DSL and my newish Uverse service and cant really tell a difference. Ping times to the same servers I always play on increased about 5-7ms. |
|
 Learux
join:2001-10-23 Valencia, CA | reply to kilometers Interesting because like I said my pings increased by~40ms.
Learux |
|
 MyDogHsFleas Premium join:2007-08-15 Austin, TX
·AT&T U-Verse
·AT&T Southwest
edit: April 17th, @06:19PM
| reply to kilometers from Austin, with Max tier (10/1.5):

|
|
  bky Premium join:2002-07-05 Austin, TX
·AT&T U-Verse
| reply to kilometers said by kilometers :I tried doing a tracert to dslreports as well and these were my results. Interleaving isn't your problem. Hop 9 is your problem (tiscali). |
|
 NormanS Premium,MVM join:2001-02-14 San Jose, CA
·Pacific Bell - SBC
| said by bky :Interleaving isn't your problem. Hop 9 is your problem (tiscali). Not really. I run 90 ms average to NYC from SJC on Fastpath DSL. Looks like normal latency from west coast to east coast. -- Norman ~Oh Lord, why have you come ~To Konnyu, with the Lion and the Drum |
|
  bky Premium join:2002-07-05 Austin, TX
·AT&T U-Verse
| Routing through germany isn't normal. However dslr seems to be getting inundated by denial of services attacks lately which probably has something to do with it, probably not the best traceroute target example. Anyway, my point was that the problems described in the OP shouldn't be blamed completely on an interleaved connection signal. |
|
  gdm Premium,MVM join:2001-06-15 Mchenry, IL clubs: | Agreed. If you tried doing a tracer today it was not the normal path to dslreports.
To try a better test google.com or yahoo.com wouldn't be bad. |
|
 NormanS Premium,MVM join:2001-02-14 San Jose, CA
·Pacific Bell - SBC
edit: April 18th, @12:46AM
| reply to bky
said by bky :Routing through germany isn't normal. And where would you place hop 6 in this trace route? Hop 6 is just 10 ms away from San José, California; and you would place it in Germany because a RIPE Whois lookup states that a German company controls that IP address?
Hop 9, in kilometers 's post, is 118 ms from California; and, by the rDNS name on that router, as well as the latency, the location is New York City, New York.
European companies are allowed to own, and operate, IP hardware within the U.S.A.
P.S. The trace route I am showing was run earlier in the day. I am not going to hammer the DSLR server with more ICMP packets while they are undergoing a DOS. But the dropped packets after my 7th hop are probably an indication of the attack. 90 ms is about right for SJC to NYC on DSL Fastpath.
-- Norman ~Oh Lord, why have you come ~To Konnyu, with the Lion and the Drum |
|