Slashdot Mirror
Maximum Latency for ISPs?
fluor2 asks: "My ISP is providing me 8mbit ADSL, and my speed is in fact 8mbit (downstream). However, we all know that there is no relation between transfer rate and latency, eg, a high transfer rate and high latency will kill your FPS games. A packet that travels through the sky and up to a satellite is bound to give high latency. Using pathping, I discovered that my ISP provides me with a latency of 22ms before my sent packets are sent out of my ISP's backbone (6 hops). I have a friend that also tried the same, and he got only 10ms before he was out of his ISP's network. I know 22ms is decent, but I still think that it's far too high if one uses IP-phones and similar. What kind of latency can we accept for a normal 8mbit ADSL connection, and isn't it about time that we get more focus on this subject?"
I've got 512/128kb and consider it to be luxury. Perth, West Oz.
Got time? Spend some of it coding or testing
An 8mbps DSL line...
Since you're one of the first folks to try out this new tech, I think you need to tell US what to expect.
How much do you pay for that thing anyways? Just to play games?
Holy shit. I have trouble putting food on the table and you're worried about your high latency times for an 8mbps connection?
What transmission scheme are they using? With CAP, you can expect lower latencies, around 10ms I think. However, most telcos are switching to DMT because I think it's more scalable. Unfortunately, DMT gives crappy latency, I've seen 60ms in some cases.
22ms latency to leave your ISP's backbone is actually quite good for DSL.
Featurewise, most cable modems are crappy, but their latency is better than DSL in most cases.
As far as VoIP goes, 300ms will still give good results. Some codecs don't play nicely with high latencies, but I've used VoIP with a 600ms latency satellite link, and it worked just fine. The latency on your TDMA or GSM phone is several hundred ms, just call another cell phone from yours and put one up to each ear and talk, there's almost a second delay.
Need Free Juniper/NetScreen Support? JuniperForum
For a little ISP, ArachNet is pretty good for connectivity.
Got time? Spend some of it coding or testing
I get 15ms to my ADSL modem, and I used to get 33ms. You are getting pretty good pings if it's still in your ISP, except about 40ms in your ISP.
I don't see whats wrong with what you are getting, maybe you are whining just a little bit too much about what you are getting.
Heck, I'd like 8mbps down on my ADSL. I'm stuck with 1.53mbps/640kbps.
Oh well. There is nothing wrong with what you get..
Free means no restrictions, ironic the FSF's GPL forces restrictions, isn't it? What's your definition of free?
isn't it about time that we get more focus on this subject?
About time, sure. If I could get anything other than no-server cable, I'd be sure to jump on your bandwagon.
Can we focus on getting decent broadband to everyone first, and THEN start worrying about 12 ms of ping time? Good god, man.
got standards? --- http://www.w3.org/
You've got an 8mbit a second ADSL connection, and you get 22ms pings? Cry me a river.
:(
Alright, yeah. I'm jealous.
RaGe
We're all just noise on the wires..
Back then, we had 33,6k modems, with 200ms pings at best, we played quake in software mode in 320x240 at 10 fps, and we were happy!
Cable modems generally ping better than DSL for whatever reason, and I'm sure even fatter dedicated lines are better as well.
On my cable modem (adelphia) I get 10-12ms for the first 8 or so hops as they are all on the adelphia servers, after that I can get as low as 20ms or even 18ms for more local stuff (I get about 23 to www.yahoo.com). I live in San Diego and this type of service is only about $35/month. On my DSL (Pacbell) I used to get 15-20ms to the first hop even, whereas i get 9-11ms now.
Verizon has 2 networks in our area, one is a T1 (fijitsu)based, the other is T3 (westell) based dsl modems.
I was on the fitjistsu on the 768/128, about a 33ms ping to the seattle bbnplanet backbone, I moved down the street, and they put in the new higherspeed network. 1500/384 and 10ms to the bbnplanet backbone.
USwest back in Spokane was about 15ms on a 768/768 cisco modem.
While I find Verizon and other telcos to be better bandwidth and ping, smaller mom and pop ISP's tend to oversell. Speakeasy was would be choice if the telco is oversold, and earthlink if ISDN is your only choice. Thou small ISP's do re-sell ISDN cheaper, and ping is good enough for multiplayer games, 20ms+. (Remember its different for each user and location!)
I'd check out dslreports and ask other people in your area. And networks change from city to city, cable/dsl/isdn/frame all depend on the routers and hop count. Plus if your ISP is a peering partner with local ISP's, they connect all major ISPs locally, thats a plus. Sometimes you notice crazy routing, like Seattle to California and back to go across town to an ISP without a local peering agreement.
Also, you call your ISP, and ask them to do a traceroute from their network to a gameserver and email it to you. I've asked this from hosting services, and who they having peering agreements with. Some will even give you a network diagram or have them posted on the site, like Verio. (Who while expensive, does seem to have good peering agreements.)
VoIP shouldn't be an issue. An additional hundreth or fiftieth of a second is not noticeable.
May we never see th
I had to fight and fight to get my ISP to take seriously my demand that the first hop be less than 50 ms or I was going to find someone else. See, I went with a provider that I thought was a local ISP who turned out to just be reselling service from...halfway across the country. So, I get the ATM link from here (Oak Park) to my CO, but I'm positive that that gateway router is in Virginia. If I wanted to give the business to my ILEC, I could probably do better, but as long as it's 50ms or less I can live with it. If I changed to Ameritech I'd probably have to give up my static IP and unblocked ability to have a small web server too.
7 November 2006: The day Americans realized corruption and incompetence weren't addressing 11 September 2001
Maybe when you're done bragging about your internet connection, you can come back down to earth. (i.e. the place where 12 ms on your first hop doesn't matter) What are you doing with this anyway? Playing games? 12 ms is not going to make any noticeable difference anyway. Get real.
I used to bulls-eye womp-rats in my pants
Ask your ADSL DCLEC if your line is configured for interleaving - many do so by default. If so, have them change that to "fast path" (aka non-interleave) and you will have lower latency at the expense of a higher error rate.
Some ISP's who use satelite also offer a Gamers plan that doesn't use the satelite and has a much higher ping. Of course since your bypassing their bulk buying satelite power you either pay more money or/and get less bandwidth.
Your post states that latency and throughput are unrelated. For TCP connections (FTP, HTTP, IMAP, POP, and many games), this is absolutely not true.
:)
The maximum possible throughput of a TCP connection is one "window" of data per round-trip time. The "window" size is essentially the amount of unacknowledged (ACK'ec) data that can be outstanding. This is often called the bandwidth-delay product, I think.
What you need to take away from this is that even if you had infinite bandwidth between you and your peer, the throughput of a single TCP connection is upper-bounded by the delay product. For example, if your window size is 32KBytes (I'm going to use 32,000 to make the numbers prettier) and the round-trip time is 100ms, then you can transmit (or receive) at most 32KB * 10 = 320KB per second. To go faster, you have to either increase the window size (which consumes more memory) or decrease the round-trip time (which is sometimes impossible, since the speed of light is a constant, or so my physicist friends claim).
A couple other points.
You're probably not capable of noticing the difference between 10ms and 20ms in terms of response time for interactive applications, including online gaming. if it were 10ms vs 100ms or 200ms, then yes, but 10ms is less than one refresh interval on your monitor, so you really can't "see" the difference.
As far as VoIP (IP telephony) and other multimedia network applications are concerned, again, you must consider the end-to-end latency (one-way delay) and/or the round-trip time, not the latency between you and some arbitrary router at your ISP.
The phone companies spec their systems (or so I've heard) such that the *round trip* latency for a domestic call is always less than or equal to 100ms. We're talking POTS here, not cell service, which experiences higher latencies.
I work on VoIP software; in an IP call (both ends are IP clients), it's very hard to keep the *one way* latency below about 100ms, if you're lucky, even if both clients are on a LAN. This is because you have to have various buffer and jitter compensation delays so that the sound quality is acceptible under somewhat adverse network conditions. In a typical call across the internet, 200ms one-way latency, IMHO, would be considered quite good.
So your 20ms intra-ISP latency (vs. the 10ms that your friend reports) is in the noise.
Oh, I should also mention, for completeness, that packet loss (or even reordering, which is more common that you may realize) can *really* hurt both TCP and VoIP (which usually uses UDP) performance/quality. This gets into some messier technical issues... basically, though, if your DSL isn't lossy, and you're getting 20ms intra-ISP latencies, you're doing as well or better than most of us.
Your friends who are running on 56k modems, who eat 200ms just to get their packets to the ISP's router on the other side of the PSTN are really going to be hurting
I'll stay off the whole soapbox of "There's wars happening in other countries and you're worried about 20ms ping times?"
But seriously, what's Slashdot going to be like 5 years from now? "Hi, I've been noticing myself with 30ms pings when deathmatching my friend who now lives in Japan. Is there any way to speed up the speed of light?"
It depends on how big your ISP is. If they immediately feed you out onto someone else's network (ie, if they're a tiny ISP or whatnot), you'll get low pings (in theory). A larger ISP (Adelphia, in my case) has like 8 hops before I go onto above.net, averaging 39 ms until I'm off adelphiacom.net. Latency on your ISP's network isn't necessarily a meaningful measurement. I'd be much more interested in ping times to certain hosts. I average ~80 ms, although this can vary hugely -- if I'm pinging sites in Asia, it'll obviously be a bit bigger.
________________________________________________
suwain_2
Hey, you think that's latency, consider people forced to use this protocol.
Some of those ISP's that offer ADSL have started to offer SDSL or VDSL. VDSL is currently very expensive in my area and only people within a short distcance from a telephone central can get it. SDSL is more flexible when it comes to max distace. Most people on SDSL get lower ping.
When I got my new connection I could either choose between 1024/512 ADSL at $85 or 1024/1024 at $140.
A bit expensive, but I get my own permanent IP, no pay per GB thing, can have my own servers etc.
And I can't complain at the latency, since many of the other users on the ISP are offices and bussiness whom almost only use their computers at office hours I get very low latency. Approx. 15 ms. to many CS-servers and the same to a backbone.
So I'm happy, but I still gaze at the connection of a friend of mine. He just got a VDSL 12500/6250 at $227. Officially, According to their User Agreement he cant't resell but the ISP is not that strict on it so he allready has 10+ customers... ;-)
Melius mori in libertate quam vivere in servitute.
22ms is decent indeed and will serve you more than well for FPS games and VoIP. For ATM networks, the maximum latency for voice is defined as 500ms. People can get by with 200ms and not know it unless theyre playing reflex games on the phone.
No its not about time we get focused here, when ISPs were over 500ms it was an issue. Below 50ms theres no issue at all unless you just WANT to have lower latency for the sake of it. And then counting hops and demarcating the bounds of your ISP gets you nowhere. If Sympatico is a reseller of Bells service, do you consider the city-wide Sympatico, Sympatico's WAN, Bell Canada, Bell global or Bell the carrier?
If the ping is below 50ms to most of the top 20 WWW sites, the ISP is good. We just have to worry about the guaranteed and the maximum burst bandwidths. I'd personally worry more about the costs nowadays and the trend among ISPS to implement rediculous monthly download caps.
"Give orange me give eat orange me eat orange give me eat orange give me you." -Nim Chimpsky
Latency-to-edge-of-network has got to be the most broken benchmark I've ever seen. If your network passes its traffic off to a different network within the same city, while my network takes it halfway around the world and passes it directly to the destination machine's network, my packets are going to be staying within my network for a long time... but they'll probably reach their destination sooner than yours.
If you're going to measure how long it takes for your packets to get somewhere, make sure you also measure where your packets are getting to.
Tarsnap: Online backups for the truly paranoid
I can't even get freaking ADSL and you are worried about a ping time that I get on a LAN.
The added latency in dsl setups can also add to irritation when web surfing as well.
Many Thanks,
Luke
You shouldn't have too much of a noticeable problem with voip with a 22ms latency. Now if the latency fluctated between 10~100 ms sporadically, you'd probably have intelligiblity issues...
Er, what in the BSD licence prevents you from profiting from your own code later?
What in the GPL prevents you from profiting from your own code later? Just the other contributors. You can only profit from your own pieces of code, which in a big project are going to be pretty close to useless by themselves. And that's totally fair.
I also reckon that it's totally fair, fine and dandy if you choose to release your own code under a licence which is open to abuse. It's your code, and I refuse to ping you for throwing it to the wolves. Now please bite your tongue (fingers?) about what I do with my code.
Got time? Spend some of it coding or testing
Your ISP may be better connected to the backbone
than your friend's ISP. What matters is not the latency
to the AS boundary, but the average latency to your
peers. Also, 8MB up with a 128k down is not going to
get much better ping time than 512k up with a 128k
down: The 128k down segment is going to dominate
your ping time (which is bidirectional).
Yes, latency sucks. It's sucking more and more as
ISPs optimize devices for b/w at the expense of latency,
although the customer base would benefit more from
decreased latency than increased b/w, so I appreciate
your effort to draw attention to the issue.
-I like my women like I like my tea: green-
Is there an easy way to run a pathping in linux? I suppose I could run traceroutes and pings manually or with a script and try to reconstruct what pathping does (according to the M$ site).
It would be nice to have a way to do it because pathping seems as useful a utility for network admin as nmap, etherape, and ethereal.
You're looking for the excellent mtr.
Believe me, there isn't anything you can do on a network in Windows that you can't do better in Linux.
May we never see th
You have a 8 mili bit per second ADSL line, and a 22 mili second latency worries you?
;p
Personly I would upgrade to a better ADSL connection, say a 8Mbps.
i live in thailand and have to be happy to ;)
;)
;)
...
get 0% packet loss
latency will be an issue here in 10 to 20 years
Tracing route to www.yahoo.akadns.net [66.218.71.92]
over a maximum of 30 hops:
1 corruption
what ADSL modem can do 8000/8000 please?
if found ADSL PCI modem doing 8000/1024
thanks!
Yes, latency and throughput are related, but
The point about TCP windows is likely bogus for this application. Most modern TCP implementations include the window scaling option, which will allow scaling to quite high data transfer rates. At these low data rates (a few megabits, or even lower for games) the windows are unlikely to cramp your style (by limiting bandwidth) One usually wants bigger windows for high volume transfers (say > 10 Mbytes/second) that you would see on a LAN.
22ms pings? Oh man, that just sucks :)
:)
I've got only a shitty 1.5 mbps adsl link, but my ping rulez yours big time! Mine are just 10ms
But seriously, 22ms pings are in fact extremely good for an ADSL line. I am extremely fortunate to be just below that, and the only reason is that these adsl lines are educational connections to a 2 gbps university network. I doubt corporate providers would be able to meet or exceed the sheer overkill in bandwidth (especially during evenings and weekends).
My advice would be: first find something that is broken because of 22ms pings, and THEN start looking at how to improve it. If it ain't broken, don't fix it.
I'll exchange my line for yours any day!
maybe it's not with the provider and the back-bone ... maybe it's your ADSL-modem?
...
...
;)
...
what's the speed on USB 1.1? if it's a modem lke that that's where to look!
if you have a network (say 100 Mbits) and you're watching a DVD sreaming from your local file-server then that is where to look
personally i prefer stuff i can stick as close as possible to RAM/Processor/HDD e.g.
PCI bus or if you have it PCI-X like the fibre-optical PCI networkcards use to get those 20'000'000'000 bit/s to the processor or from the HDD
this why i got a PCI ADSL modem
USB is good to connect your mobile phone for a recharge
WI-FI sucks. bluet00th rulez!
i got the ASUS A7N8X Deluxe motherboard.
v er view.htm"
;o
url:"www.asus.com/products/mb/socketa/a7n8x-d/o
it's got 2 100mbit/s NICs.
i use both nics on all 8 maschines.
first network has TCP/IP only.
second network has TCP/IP with filesharing to SAMBA (i would have prefered not running TCP/IP on the second network at all, but LINUX and SAMBA don't understand NetBEUI. NetBEUI because it is none-routable->more security!)
second network also runs IPX (Quake 3 Arena) and NetBEUI (print-server).
sorry no apple-talk
i had to get twice as many hubs, but now gaming, file-share,etc. on the local network does not interfere with the access time to the internet!
PCI ADLS modem under LINUx(NAT,SAMBA,SQUID...)
oh, and yeah, one IntelPRO networkcard (49US$)(1'000'000'000 bit/s) to the WAN
ASUS rocks, horray to Nvidia, faster ATHLONs please!
latency was a non-issue until it reached 10ms. above that, a ticket was opened.
22 ms is pretty good. However, as everything, it depends. For example, how large is your ISPs network and how close does it get you to the final location you are interested in? For example my cable ISP has a larger network. If I try to contact a server a few states away, it uses my ISPs lines for most of the trip.
If you have a service level agreement, it usually specifies 100ms as maximum round trip time within the ISPs network. I guess they pick this rather high number as it usually is fast enough and shorter times are a bit hard to measure.
Also: Dont necesserily trust tools that use ICMP packets to measure roundtrip times. Some ISPs implement QOS rules that give ICMP a lower priority. Try UDP, or if you use TCP make sure you set the TOS flags for low latency.
---- join dshield.org Distributed Intrusion Detec
I get a round-trip time of about 12-13ms between my parents ADSL connection (Telia, 512kbps) and the last Telia router. I get about 16-17ms pinging my machine on a Swedish (sunet) campus network. That's 16 hops away...
22ms is definitely good. Especially leaving your isp.
:)
Dsl speeds won't affect latency too much. I know it's not supposed to, but it does.
A 1184/160 dsl will ping around 15ms to its gateway
A 1728/384 dsl will ping around 11ms to its gateway
A 3488/800 dsl will ping around 7ms to its gateway
I have a feeling its more related to the upstream speeds than the downstream. An 8MB dsl has an 800 upstream maximum so the pings will most likely be the same as a 3MB dsl. Isp's can have different upstream speeds for all downstream speeds.
All these speeds are assuming your isp doesn't put you on interleaved channel which gives you 55ms+ ping to gateway. And if you have a decent isp further hops will only increase by 1-3ms a hop.
As for my experience with several cable providers ping to gateway is 6ms. But all cable providers i've had go to hell during peak periods. I know this is not the case for everyone. When i originally had cable 5 years ago i had no problems, then again no one else had cable... And i enjoyed 20ms in game.
10ms might not sound like much but to a gamer it does
The other thing is, that you shold really only be interested in end-to-end RTT, not the individual hops. For example, if there's cisco 4xxx series switches with SUP-3's out there, your icmp/traceroute IP packets gets processed in the processor card, not on the interface, causing an 10ms more latency to pings/traceroutes. The actual IP packets of your connection get forwarded by the interface so no latency there.
Your not always going to get the same FPS on every game, in every server. Lantency is dependent on the distance between you and the server.
I have speakeasy sdsl [shameless plug] and have a 10ms time to google.
For games, it's ~50ms for anything on the west coast, and ~10-30ms for anything on speakeasy's network.
I just did a quick traceroute from the webserver here where I work through a 1.5Mmbs or so ADSL w/ Qwest (up and downstream pretty much the same) to my home computer (standard Comcast Cable). To my suprise, from the router here to its first hop was +38ms... but from my home connections gateway to my home computer is only +10ms. Yes, I did it a few times just to make sure it wasn't a fluke... but it was pretty consistent around that.
This is a very expensive ADSL line that we have going here... I would think that it would perform a little better than my home connection.
I heard some MCI execs were hoping for a latency of 5 years + good behavior.
Sheesh, evil *and* a jerk. -- Jade
I had myself a cable internet conenction this year.
I liked the speed, frequently between 2 and 3 mbps. But, I complained about the lag. My ISP didn't get a complaint phonecall until my GATEWAY became at least 150ms away (instead of the normal 75ms), let alone the internet...
I found that average internet ping time was 600ms. this was a problem. I was used to a fast cable provider which could offer internet ping times in the 150-250ms range. (and gateway pings in the 20-40ms)
So please, stop complaining that your internet access is only instantaneous.
If you aren't happy with your DSL connection, I have a very nice 14.4 modem that I would be happy to trade you.
What would be a good tool for me to use in 'nix to figure out the basic latency of my connection? I suppose I could just ping out a well-known host, but that would also involve the latency at their end?
The sadylite will alway be faster ...it runs downhill to get to you.
karma 10 to the negative 10000000000000000000000
BTW, I can't understand why anyone would bother modding down the AC who replied to the parent. Surely you can find better places to spend mod points? And if not, give him a +1 Funny or explain what you have against Python misquotes.
Got time? Spend some of it coding or testing
% sudo mtr www.yahoo.com
Packets Pings
Hostname %Loss Rcv Snt Last Best Avg Worst
1. myhost 0% 9 9 0 0 0 2
2. sfo1.dsl.speakeasy.net 0% 9 9 14 13 14 15
3. border5.g3-4.speakeasy-29.sfo.pnap. 0% 9 9 14 13 18 46
4. core4.ge2-0-bbnet2.sfo.pnap.net 0% 9 9 15 14 51 207
5. so-1-3-0.0.ar4.sfo1.gblx.net 0% 9 9 17 14 21 62
6. pos1-1.core1.SanFrancisco1.Level3.n 0% 9 9 14 14 16 18
7. so-4-0-0.mp2.SanFrancisco1.Level3.n 0% 9 9 16 14 20 54
8. so-2-0-0.mp2.SanJose1.Level3.net 0% 9 9 18 15 17 18
9. gige10-1.ipcolo3.SanJose1.Level3.ne 0% 9 9 16 16 17 19
10. unknown.Level3.net 0% 9 9 17 17 17 19
11. w7.scd.yahoo.com 0% 9 9 18 15 17 19
The first thing that makes a VoIP call bothersome as latency rises is the echo. If the person you talk to has a good echo canceller, you will be OK up to about 150msec. After that, you start to attribute delays in each others' reactions to emotions, usually reluctance, reulting in anger.
Everyone here seems to be suffering from the same delusion that a fatter pipe means a faster line.
Boy have a I got bridge to sell you! ( and some terrific coastline property too!)
No matter how fast you can download data, all the signaling involved has to obey the laws of physics.
Expecting to get less than 20ms of latency getting from your pc to your ISP's connection to the NET is extremely unrealistic (the average T1 introduces 20ms of latency) just because of the number of signal + data processing devices involved.
Sure having a fatter pipe MAY mean that the time it takes for your packet to traverse the link is reduced, but we are still talking something on the order of fractions of a millisecond.
Considering that the average human takes 1/2 second to process visual changes in their environment, and another 7/10ths of a second to physically respond to that processing, it is highly doubtful that changing ISP's to save 10ms of latency will actually improve your fragging experience.
The reality of this is that your average path is something like this:
gaming rig
hub/switch
router/firewall
cable/dsl modem
local carrier head-end
local carrier agregate routing device
local carrier backbone
Local carrier border routing device
ISP agregate routing device
ISP core routing device
ISP border routing device
NSP border routing device
NSP core routing device
NSP backbone
NSP core routing device
NSP border routing device
ISP border routing device
ISP core routing device
--then if you are lucky--
ISP colo routing device
ISP colo switch
Gaming server
Now regardless of whether or not your traceroute actually shows those first 10 hops just to get to the NET, Your packets do make that whole trip, (just because those devices don't respond to ICMP requests doesn't mean they aren't there)
Now considering you have an average of ~4ms through each processing device, that gives you upto 40ms of latency just getting to the Net (not including transit time over the cables).
Heck I have customers running dedicated OC3's for their server farms and they are happy with their 10-20ms transit latency. Mainly because they know that even across OC192's that their coast to coast latency is going to average 80ms.
That is just using the good old speed of light to calculate transit time and an assumed average of ~4ms processing time.
From an engineering point of view, it depends upon your requirements, but for public access internet providers, lowish latency is important.
We use the term "bandwidth delay product" rather than "bandwidth" as this refects the combination of speed and latency.
traceroute to slashdot.org (66.35.250.150), 30 hops max, 38 byte packets
1 my.gateway (216.xxx.xxx.xxx) 0.690 ms 19.335 ms 0.404 ms
2 some.machine.at.savvis.net (216.xxx.xxx.xxx) 1.929 ms 1.903
ms 2.108 ms
3 500.POS2-1.GW4.ATL3.alter.net (157.130.81.41) 2.266 ms 2.175 ms 2.007 ms
4 147.at-1-0-0.XL3.ATL1.ALTER.NET (152.63.81.50) 2.449 ms 2.797 ms 2.680 ms 5 0.so-7-0-0.XL3.ATL5.ALTER.NET (152.63.85.190) 3.104 ms 3.372 ms 3.207 ms
6 193.ATM6-0.BR1.ATL5.ALTER.NET (152.63.80.113) 3.139 ms 3.127 ms 3.056 ms
7 204.255.168.74 (204.255.168.74) 3.841 ms 3.916 ms 3.843 ms
8 agr3-loopback.Atlanta.cw.net (208.172.66.103) 4.301 ms 4.316 ms 4.227 ms
9 dcr1-so-0-2-0.Atlanta.cw.net (208.172.75.9) 4.243 ms 4.106 ms 4.649 ms
10 dcr2-loopback.SantaClara.cw.net (208.172.146.100) 69.385 ms 69.408 ms 69.319 ms
11 bhr1-pos-0-0.SantaClarasc8.cw.net (208.172.156.198) 67.176 ms 66.858 ms 66.888 ms
12 csr1-ve240.SantaClarasc8.cw.net (66.35.194.34) 67.468 ms 67.247 ms 67.320 ms
13 66.35.212.174 (66.35.212.174) 71.527 ms 71.836 ms 71.561 ms
14 some.machine.at.slashdot
Are you kidding? 22ms is great for voice chat. Think about it ... You're talking about 1/50th of a second here. You really can't tell, trust me.
... there's lots of work in this area.
:-)
The main culprit in VoIP latency is really jitter. That's basically packets that arrive out of order or don't get there and need to be resent. If you've got a lot of jitter it drives up the latency to compensate (the codec needs time to reassemble them in the right order). It's better often just to drop missed packets
Jitter is usually caused by congestion. So, as long as you aren't saturating the link, you should be fine with 22ms
simon
home page
While some may call this nitpicky, I constantly am on the receiving end of many many otherwise informed folks that think that ICMP PING is an accurate test of network performance.
It's not.
PING was intended for reachability checking, and as a secondary feature, response time.
The ICMP part of most IP stacks often has the lowest priority to receive CPU time in a lot of IP stack implementations. When you PING, the return packet back to you is at the mercy of the resources of the system writing, generating, and spitting that packet back across to you, and if its' CPU is busy, you're going to get a high latency time that is not accurate. For those of you with Cisco gear or really anyone's network kit, try pinging a router when it's busy doing routing calculations (like OSPF LSA expiry or an Area-0 event).
The only accurate way to test the end-to-end throughput, goodput, latency, and jitter, is to stream packets from one end to another and measure across time. See the NTOOLS or HPING projects on FreshMeat, etc.
This is a simplification but...
The time it takes for a packet to reach its destination is the (A) time it takes to transmit it plus the (B) propagation delay.
A = packet size / transmission speed (your bandwidth)
B = speed of the signal on the physical medium