dslreports logo
site
 
    All Forums Hot Topics Gallery
spc

spacer




how-to block ads


Search Topic:
uniqs
2341
share rss forum feed


JigglyWiggly

join:2009-07-12
Pleasanton, CA
reply to hhahn

Re: [Praise] I cant believe how low my latency is on Comcast

aw no fair you guys cheat


EG
The wings of love
Premium
join:2006-11-18
Union, NJ
kudos:10
reply to hhahn
said by hhahn:

Ping is an extremely poor metric to gauge the quality of a connection, without a baseline. While a lower ping is better, it doesn't mean anything alone. As others have mentioned, it has a lot to do with the distance from the speedtest server. It's not like a magical number that rates the quality of your connection. But, it doesn't just deal with the geographical distance from a speedtest server, since it includes the number of hops for that test specifically. What I'm basically trying to say is that you can't judge a connection by ping alone. I have both Comcast and FiOS. My Comcast connection has a 12ms ping to the Philadelphia speedtest server, while my FiOS connection has a 25ms ping to the speedtest server. That doesn't mean Comcast is superior; it just means the route to the Comcast speedtest server is shorter because you are testing from within the same network. A route to a different server (like a website) may be completely different and yield equal or opposite results.

However, ping is important to know as a baseline. You want to know if it varies at all between tests to the same server (jitter), or if it varies depending on the time of day. This can help a great deal in troubleshooting. For example, if you notice your ping is normally 5ms but now it's 200ms, there may be a problem with either your connection or a bad route. Alone, ping means nothing.

Today widespread usage of ICMP based ping packet de-prioritization is also a large factor of diminishing the efficacy of ping as a troubleshooting tool.


MysticGogeta
The Robot Devil
Premium
join:2005-03-14
Katy, TX
reply to nine9s2
Pretty solid for me at my new place



--
Team Discovery-Join the fight

n_w95482
Premium
join:2005-08-03
Ukiah, CA
reply to RR Conductor
said by RR Conductor:

WOW, they must be offering the new Tiers here now (Netcool said it'd be soon, since we have the new provisioning now). I probably don't want to know what that costs though. lol Blast! 20/4 at 72 bucks a month is enough.

I just checked on the main Comcast site, still the same old listings. I'm still on 16/2, but business has uncapped PB. 22/5 is $10/month more than 16/2, which is one of the reasons why I've been so anxious for them to roll it out here .
--
KI6RIT


telcodad
Premium
join:2011-09-16
Lincroft, NJ
kudos:15
reply to nine9s2
FYI - An article on the GigaOM site today about Latency issues for residential HSI:

Low-latency networks aren't just for Wall Street anymore
By Stacey Higginbotham, GigaOM - April 6, 2012
»gigaom.com/broadband/low-latency···anymore/

The intro:

"When it comes to broadband we spend a lot of time talking about how much data we can transfer in a set amount of time (measured in Mbps). But what about latency? Applications from video chat to gaming need networks where packets get to their destination quickly, which requires a different type of network speed.

Plus if we're going to stream everything from video to music in addition to our latency-sensitive apps, we’re going to have to add some metrics to our definition of broadband and some intelligence to our home networks. The question is how carriers can respond to both the threat and the opportunity this development offers and what other companies might play in this space."

JPnATL

join:2011-11-16
Bethlehem, GA
reply to nine9s2
50 miles to the server and back it should be low.

voipnpots

join:2011-10-13
USA
reply to nine9s2
I usually get 9ms ping. Here are my new Performance speeds that were upgraded (I live in Florida)!




jimi419
Dadof4

join:2002-03-14
Round Lake, IL
reply to nine9s2
thats on comcraps own network im on the blast tier 30/2-5?? totally crap speeds imo heres a test to chicago leaving the comcast network Download Speed: 4.428 Mbps (4428 kb/sec transfer rate)
Upload Speed: 4.571 Mbps (4571 kb/sec transfer rate)
Friday, April 06, 2012 11:20:32 AM that is using a different test site
and im using a surfboard 5100 with a netgear WNR1000 wireless/wired router im wired in wireless is for the lappy and ps3


SpottedCat

join:2004-06-27
Miami, FL
reply to nine9s2
Ping is indeed fairly good with Comcast. Unfortunately they seem to suffer from jitter, which can mean your round trip time to your default gateway (which is only one hop away!) can vary between 5ms and over 100ms.

Check out this ping graph for the past 32 hours. Every five minutes, ten pings are sent, and the lowest time is graphed in blue and the highest in green. Look at how many green spikes there are!


Ping bloat

@mycingular.net
said by SpottedCat:

Check out this ping graph for the past 32 hours. Every five minutes, ten pings are sent, and the lowest time is graphed in blue and the highest in green. Look at how many green spikes there are!

The main reason this data is invalid is because so many people ping blast infrastructure causing ISPs to rate limit ICMP. Your conclusions are wrong and the false reporting is due to your (and others) arm chair network monitoring.


SpottedCat

join:2004-06-27
Miami, FL
If ICMP were being rate limited, the result would be dropped packets, not delayed packets. Delaying ICMP accomplishes nothing; it's not like TCP where delaying a packet causes window sizes to shrink, lowering the transfer speed.

I *do* know what I'm talking about. At least for the most part.


28619103
Premium
join:2009-03-01
21435
Measuring your latency and jitter is best to do from a known (managed by you) destination/end point and not ISP infrastructure.

bman212121

join:2005-06-09

1 edit
reply to SpottedCat
Hmm... I wonder if something is configured differently in your area. You might even want to check the connection without a router. We have very consistent response times although I will say ping times have gone up since I switched to a D3 modem by about 2ms.




Running a simple ping for a minute from a command prompt shows min 6, max 22, and average of 8.

I'm not sure if I would agree with how the testing is being done though. An average would definitely be a good thing to have when looking at the numbers. If you have 9 packets that make it there in 10ms and one that is at 100, it doesn't necessarily equate to bad performance.