so youre saying that you got a lower ping to servers near you? but seriously, i think its mostly your routing, here is what i get (with my rather crappy connection):

Antwort von 91.198.152.137: Zeit=18ms
Antwort von 91.198.152.137: Zeit=16ms
Antwort von 91.198.152.137: Zeit=15ms
Antwort von 91.198.152.137: Zeit=17ms

how is it my routing when i have 6 ms until i hit i3d router and ql server? i've gone from 29 ping to this over the years in ql now for german servers. i had this exact same route a few months ago when i was still pinging (again after a long period of playing with higher ping) a steady 38 (and i had 38 for like half a year). now i get random jump ups of 10-25 ms because of shitty i3d.net just randomly depending on the day.

well i ping below 20ms and you experience 10-25ms from a node, or am i missing something? dont know much about network stuff, maybe i cant compare a traceroute and a ping like that

Well its starting to have been some time ago since this changed around, but like you I used to have good ping to germany. Which is basicly next door to me.
Used to have between ping 28-32.

Here is the danish side of things, I am about 50km from Germany.

This is from work with ISP: Nianet

10 3 ms 3 ms 3 ms 10.200.254.246
11 3 ms 3 ms 3 ms 10.2.8.6
12 4 ms 3 ms 3 ms 193.163.229.17
13 6 ms 6 ms 12 ms ge1-16.cr1.taas.gc-net.eu [77.243.32.201]
14 6 ms 6 ms 6 ms tengigabitethernet9-1.ar3.cph1.gblx.net [208.48.13.229]
15 35 ms 33 ms 39 ms interactive-3d-b-v.tengigabitethernet7-2.ar5.fra3.gblx.net [159.63.22.134]
16 33 ms 33 ms 33 ms quakelive-hosted-by.i3d.net [91.198.152.137]


This is from home with ISP: Telenor

2 7 ms 7 ms 7 ms lo0-1.bb03-jd395.ip.telenor.dk [94.144.1.139]
3 13 ms 13 ms 13 ms ae0-0.dr01-jd395.ip.telenor.dk [94.144.2.50]
4 10 ms 10 ms 10 ms cybercity-2.ti.telenor.net [148.122.9.26]
5 10 ms 10 ms 11 ms ti3004b300-ae2-0.ti.telenor.net [148.122.9.25]
6 21 ms 21 ms 21 ms ti3004c310-ae0-0.ti.telenor.net [146.172.105.41]
7 21 ms 24 ms 49 ms ti3001c310-ae6-0.ti.telenor.net [146.172.99.193]
8 21 ms 21 ms 21 ms ti3001b400-ae0-0.ti.telenor.net [146.172.105.25]
9 51 ms 49 ms 51 ms netnod-ix-ge-b-sth-1500.i3d.net [194.68.128.206] - Same place that gives frs increased responsetime.
10 46 ms 46 ms 45 ms quakelive-hosted-by.i3d.net [91.198.152.137]



edit:
This is something very intresting that I just fell over, I did a traceroute from home to our dear amsterdam quakeservers: 213.163.65.211

5 10 ms 10 ms 10 ms ti3004b300-ae2-0.ti.telenor.net [148.122.9.25]
6 21 ms 20 ms 20 ms ti3004c310-ae0-0.ti.telenor.net [146.172.105.41]
7 21 ms 20 ms 21 ms ti3001c310-ae6-0.ti.telenor.net [146.172.99.193]
8 21 ms 21 ms 20 ms ti3001b400-ae0-0.ti.telenor.net [146.172.105.25]
9 48 ms 48 ms 48 ms netnod-ix-ge-b-sth-1500.i3d.net [194.68.128.206] - This is the exact same hop that gives me bad ping to germany :(
10 54 ms 51 ms 51 ms fr0-cr0.smartdc.i3d.net [109.200.218.77]
11 55 ms 51 ms 50 ms 30ge-1-1.br2.smartdc.i3d.net [188.122.95.92]
12 50 ms 49 ms 49 ms hosted-by.i3d.net [213.163.65.211]

yea its simply an extremely shitty router used by the i3d.net ql hosters

i live in frankfurt, used to ping 9ms to the server now it's 46-48 again. started about a year ago, around feb/march 2012 :(

ping 91.198.152.137

PING 91.198.152.137 (91.198.152.137) 56(84) bytes of data.
64 bytes from 91.198.152.137: icmp_req=1 ttl=51 time=43.8 ms

traceroute 91.198.152.137

traceroute to 91.198.152.137 (91.198.152.137), 30 hops max, 60 byte packets
1 easy.box (192.168.2.1) 0.371 ms 0.630 ms 0.757 ms
2 dslb-094-219-096-001.pools.arcor-ip.net (94.219.96.1) 13.766 ms 15.100 ms 17.712 ms
3 92.79.244.189 (92.79.244.189) 20.623 ms 20.840 ms 22.309 ms
4 92.79.212.65 (92.79.212.65) 80.211 ms 80.286 ms 80.391 ms
5 85.205.15.34 (85.205.15.34) 68.063 ms 68.222 ms 68.338 ms
6 ae29-0.mad-001-score-1-re1.interoute.net (195.81.199.45) 62.214 ms 44.357 ms 44.181 ms
7 ae3-0.par-gar-score-1-re0.interoute.net (89.202.161.22) 49.798 ms 50.764 ms 51.884 ms
8 ae0-0.par-gar-score-2-re1.interoute.net (212.23.42.26) 51.642 ms 52.630 ms 54.007 ms
9 ae1-0.fra-006-score-2-re0.interoute.net (212.23.42.30) 56.666 ms 48.413 ms 47.655 ms
10 10ge.de-cix.i3d.net (80.81.194.107) 44.622 ms 55.857 ms 56.156 ms
11 * * *
12 * * *
13 * * *
14 * * *
15 * * *
16 * * *
17 * * *
18 * * *
19 * * *
20 * * *
21 * * *
22 * * *
23 * * *
24 * * *
25 * * *
26 * * *
27 * * *
28 * * *
29 * * *
30 * * *

your isp is clearly fucked up :/

you are already at 50-60 before you even exit your own isp's backbone lol (some hops are even 80 ping wut?)

iplocation.net is your friend.

"2 dslb-094-219-096-001.pools.arcor-ip.net (94.219.96.1) 13.766 ms 15.100 ms 17.712 ms" first hop out of your apartment:

IP Address Country Region City ISP
94.219.96.1 Germany Saarland Sulzbach Arcor Ag

"3 92.79.244.189 (92.79.244.189) 20.623 ms 20.840 ms 22.309 ms"

still in sulzbahk but it you moved to Vodafone D2 Gmbh now

"4 92.79.212.65 (92.79.212.65) 80.211 ms 80.286 ms 80.391 ms"

still in sulzbach, Arcor Ag & Co.kg , back to arcor? nice ms increase there..

"5 85.205.15.34 (85.205.15.34) 68.063 ms 68.222 ms 68.338 ms"

vodafone ip, location unknown. unneccesary tourist location much?

"6 ae29-0.mad-001-score-1-re1.interoute.net (195.81.199.45) 62.214 ms 44.357 ms 44.181 ms"

IP Address Country Region City ISP
195.81.199.45 Spain - - Interoute Communications Limited

interoute is one of the firms who handle isp to isp connections. SPAIN??? the mad in there probably means madrid

"7 ae3-0.par-gar-score-1-re0.interoute.net (89.202.161.22) 49.798 ms 50.764 ms 51.884 ms"

still in madrid

"8 ae0-0.par-gar-score-2-re1.interoute.net (212.23.42.26) 51.642 ms 52.630 ms 54.007 ms"

oh hey now your are in france, paris ahahah

"9 ae1-0.fra-006-score-2-re0.interoute.net (212.23.42.30) 56.666 ms 48.413 ms 47.655 ms"

still in france. first had par affix so you arrived in paris, this one has fra so you moved to the part of the backbone that handles the exits from france here.

last hop is obviously DE.

in short, your isp travels all over europe to get to the server in the same city, magic of routing right there. distance has sometimes nothing to do with ping, yours is the perfect example.

if you used the vpn i posted below (it has like 3-4 de servers) you would get 20 ping for sure. maybe less?

thanks for your post, i will respond properly once i have time. now just this, my "isp" is arcor but in germany every adsl line is basically still part of the german telekom network, in other words, it's not the isp (or just me) but almost everyone on adsl in .de (afaik).

adsl providers rent their lines from the telekom, which used to be state sponsored but got privatized (and the market openend) a few years ago.

if that were so everyone would ping high to the ql server. but thats not how routing works. everyone who has adsl.

it goes alot deeper then that concerning routing. some isp's make their own backbones and cable system (phsical), others rent them from other isp's. and alot more other details that dont really matter in your case.

the part remains that your default isp routes you all over europe to reach the i3d network. you will either have to change isp or use a vpn, and like I said connecting to a DE vpn server will do wonders for you.

i agree, just this one thing: i'm under the impression (might be wrong, doubt it) that adsl in .de means using the old telekom backbones. i'm also rather sure that what i'm experiencing is something i share with quite a lot of german players currently.

calling my isp might help as i remember i did this 8-10 years ago when i would constantly hassle them and it would actually work wonders :)

yes, a vpn would work. thing is, i already pay for one and have two others as backup, none of them are .de for obvious reasons. i'd hate to have to pay for one just to play quake but will look into it.

also, the routing did change with some of the quakelive updates. like the second i'd login after one my ping had changed. that doesn't change that my routing is the fault, not id, still... :C

thanks again for your lenghty explanation where my isp routes me on the way to the frankfurt server, this will hopefully help some when contacting them.

surprising that none of the 3 vpn services you use dont give a DE server. it is a rather popular location to give as far as I saw. makes sense too given germany's location

i just realized i could probably use my uni account as a vpn. will try that instead first :)

so how did it work out?

my uni abandoned their vpns two years ago :( will see into incloak now.

wut. they dont use anything.

when you connect somewhere its your isp who picks the route to the server, it has nothing to do with the final destination.

if there was any problem with the final destination itself everyone would have lag spikes, or high ping etc no matter what, but you see low ping players every day still with stable connections.

why do people always seem to blame the server..

look, its all routing, you and amokchen even posted the proof for it. amokchen's nianet uses a different backbone

14 6 ms 6 ms 6 ms tengigabitethernet9-1.ar3.cph1.gblx.net [208.48.13.229]
15 35 ms 33 ms 39 ms interactive-3d-b-v.tengigabitethernet7-2.ar5.fra3.gblx.net [159.63.22.134]

clearly superior routing here. this hop is still strickly his isp territory, and the next hop is the server itself which has no delay at all.

moving on to telenor:

8 21 ms 21 ms 21 ms ti3001b400-ae0-0.ti.telenor.net [146.172.105.25]
9 51 ms 49 ms 51 ms netnod-ix-ge-b-sth-1500.i3d.net [194.68.128.206]

this hop is the one he moves from denmark to germany, this is where the problem is basically. it is purely the relation of his telenor isp and the i3d hosting (or it might be even a isp) backbone thats causing problems.

of course you cannot do much with your own isp to change this, you cannot call them and ask to be routed differently. you CAN however change your routing by using a vpn. i get -30 (or -50 if normal isp routing is really horrid that day) by using one.

my normal isp routing goes all over the place within the country, goes to hungary, goes to france, or sometimes germany after, then within germany it goes to the ql servers backbone and finally the server itself.

by using a vpn (istanbul server) i cut through all the bs hops within my country, it goes directly to istanbul, and uses the country exit from there directly to germany on a completely different backbone (isp germany backbone would be 80 ping already in DE, to this its like 60) and from that its +10 to the ql server. 70 vpn vs 100-120 isp default routing. this routing is completely unrelated to my isp, it uses whatever isp the vpn server uses at the istanbul location.

now ofc you guys already face alot less hops then i do from your countries. but you still have potential to cut back from some hops, or you might even end up on a different exit route from your country to DE, specifically amokchen might end up with his work route for example (nianet), you never know if you dont try.

there are alot of vpn services out there. I recommend www.incloak.com (real free trial too, thats rare). it has quite alot of servers to try

well as you can see from my tracert the problem really isnt in finland or with my isp, but a shitty i3d.net router in sweden, which is the same host as the ql server host. its close to impossible to have an even more direct route to the server, only thing you can hope for is to have a different route that is just as direct but without having a shitty router somewhere there. and i tried some proxies that snapcase provided, that was before this shit though so they might actually be better now

yours is the only example that actually uses a i3d.net backbone location in this thread, and yes your ping is high to it. fortunately you can probably by pass it with a vpn, there is nothing else you can do.

you are probably correct btw, that i3d shouldnt ping so high from finland. the i3d part in germany itself is fine though.

i dont know why ppl have such bad routing to the i3d network in general, it has all to do with agreements made between hosting companies and ofc we cant check those easily.

addition, here is what I mean with vpn specifically:

my isp routes badly to i3d germany badly ( not to mention retarded tourist hops before that). so with the vpn service I by pass the unneccesary hops, connect to a different backbone in germany first, and that one has only a small delay to the i3d germany backbone ( 10 ping hop vs the 40-50 ping hop i get directly to it from my isp route) this -30 -40 decrease is key.

Quality post. Mega pluses.
Can you describe how to find/test/acquire a VPN connection? I bet most people here (myself included) wouldn't even know where to start.
Is proxy an option here?

well there are plenty of websites you can easily find with google, just write vpn and it will already give you many options. I recommend one here but there are in actuality 50 or more of em!

the problem is that nearly all have no free trial, because most people use vpn's for the anonymosity they provide and not gaming it doesnt matter much for them and they just buy one that provides the most servers.

I got lucky with my choice with incloak, it is also the only I noticed that allowed you to choose servers for UDP performance (ping) the server I need is istanbul and it is also pretty stable. some vpn servers I tried in other countries were very poor performance wise sometimes (not enough bandwitdh etc)

you can try this one because it has alot of servers all over the world (1 day free trial too, no payment at all, others charge you for a month and you apply for a refund if you dont want to use it after trying). America canada russia norway turkey, you name it.

you are in the USA (yeah i still remember you :)), and I know routing was always horrible there, you can try this and look for other options as well, many vpn providers have america servers.

Thx man :)

I wonder if there are already people who use VPN servers to connect to QL in USA and could post here for others.

I'll check when I get home, got two different ISP's I can try from.

But I've noticed I'm pinging like 30'ish to Germany in game, and 15 to Swedish servers, and when I played a couple of years ago, it was the other way around.

First ISP, Parknet 50/50 - usually running 100/75 or so.

1 <1 ms <1 ms <1 ms 192.168.1.1
2 3 ms 1 ms 1 ms 10.19.28.1
3 4 ms <1 ms <1 ms 192.168.19.1
4 <1 ms <1 ms <1 ms 192.168.0.4
5 1 ms 1 ms 1 ms br1-taastrup-xchg.parknet.dk [80.71.128.225]
6 1 ms 1 ms 1 ms 10ge6-2.cr2.taas.gc-net.eu [94.101.208.189]
7 1 ms 11 ms <1 ms 10ge6-3.cr1.taas.gc-net.eu [77.243.32.165]
8 1 ms 1 ms 1 ms tengigabitethernet9-1.ar3.cph1.gblx.net [208.48.13.229]
9 29 ms 37 ms 28 ms interactive-3d-b-v.tengigabitethernet7-2.ar5.fra3.gblx.net [159.63.22.134]
10 28 ms 28 ms 28 ms quakelive-hosted-by.i3d.net [91.198.152.137]

Second ISP, YouSee 100/10

Coming right up!

here is turkish(asia part) side of the story (2700km away from Frankfurt):

clearly shows that their routing is fine for me (10mbit fiber internet)

really my issue here has nothing to do with routing as mentioned in the thread, i have an almost direct route to the server. the issue is i3d.net using some microwave router in sweden that is shitty as fuck

I would really hope the router is not placed in sweden, that would mean I get routed to sweden before I get to germany as we have raised response on the same hop.

yes that node (194.68.128.206) is in sweden, the i3d sweden router is causing you guys problems.

you need to find a way to bypass that as plenty of people do fine to the i3d germany part.

Jesus that sucks, it is exactly the same for amsterdam quakeservers. I get routed trough sweden then.

Germany hop: netnod-ix-ge-b-sth-1500.i3d.net [194.68.128.206] <---------- placement sweden :((((
Netherlands hop: netnod-ix-ge-b-sth-1500.i3d.net [194.68.128.206] <------ placement sweden :((((


Google map placement

6 6 ms 6 ms 6 ms tur1-tr1.dnaip.fi [62.78.107.96]
7 6 ms 6 ms 6 ms tuk1-tr1.dnaip.fi [62.78.107.51]
8 46 ms 48 ms 48 ms netnod-ix-ge-a-sth-1500.i3d.net [194.68.123.206]
9 54 ms 54 ms 54 ms 30ge.cr1.smartdc.i3d.net [109.200.218.86]
10 54 ms 62 ms 54 ms 30ge-1-2.br2.smartdc.i3d.net [188.122.95.84]
11 54 ms 54 ms 54 ms hosted-by.i3d.net [213.163.65.211]

here's mine to amsterdam, different ip but its still in sweden and same hoster.. zzzz

not that it matters, ql is only played on german servers anyway

yep, like i said you have to find a way to bypass that hop. also interesting to see that the nl servers are also i3d.net, people always saying nl servers are superior (with no knowledge about routing ofc) and its the same company (and same backbone) ahaha.

this is not the first time I see this, DCR had the same problem as well (bad swedish hop). the germany part is fine though.

I would be more intrested in finding out who could fix bad routing like this, if its common for different ISPs and countries.
Surely its a matter of collecting enough info and getting the right people on the problem.
I doubt this is only a problem for quakers, but we might just be more aware of it.

you cannot fix it directly, unless your isp has other options and you reach someone who can change your routing for your ip (unlikely that your isp has different options, and hard to get ahold of someone who can even do it, if possible)

its of course different for every country, you can see that I researched mine extensively from this thread. Each country is again different but the principles are the same that each isp can be different or the same depending on how they arrange their infrastructure (self constructed or rented from other isp).

the solution is always the same though if you cant change your isp routing. vpn or different isp.

or get the guys in charge of the router in sweden to fix what's wrong with it or get a new one. what else could explain it being so bad? i should be pinging ~10 to sweden

unfortunately i dont think there is anything actually wrong with it in their eyes. these guys are not gamers so all they care is that they have a stable running server that meet their requirements of running X amount of data per day without having bottlenecks. usually ping is never part of their schedule unless it reaches alarming amounts (lets say above 500? or heavy spiking) and that is usually in a case of bottleneck anyway.

but it never hurts to try!

the emotions your post gives me, years ago I remember being on the phone because of a sudden ping problem trying to reach a tech guy who didn't tell me that a ping of 100 is very good and "no problem at all."

Well, in 2013 there actually even are players saying that 100 ping is very good and "no problem at all".

Yeah, I would still have good ping to Germany if that exact hop dindt increase my ping so much.
Even with the weird routing: Denmark at German border -> Sweden -> Germany

You should take in account that many things are hidden in traceroute from user's perspective and what you see as one hop can be a huge transit through many ISP networks. It doesn't have to be a problem with one particular router.

fucking sweden. thday tooook ourr jobss

its in stockholm. i ping ~10 to swedish ql servers and aroud 8 in tracert to sweden ...

I know nl players that have awesome routing(i think they use zippo or some shit) to nl/de(they both use i3d if i am not mistaken).
Change your isp or use a tunnel/vpn.

just know that i avoided i3d.net servers in BFBC2, was pointless to play on them.

Whilst very annoying, it isn't easy to pin point the culprit here.
When you state "shitty router", the router isn't shitty itself, it's just that it's not a node part of the shortest routing path.

It could be the server provider that chose a bad provider and is connected to a not well-served backbone.
It could also be the ISP, but not directly.
Keep in mind that an ISP isn't owner of most of the links it uses.
Most of the peering is outsourced, to external providers (proxad,colt,verizon,level3, etc)
Each provider/country has a limited sets of AS (autonomous systems), and they advertise routes to each other.
When you buy peering to an AS, you don't necessarily know how's the routing inside.

The day the provider decides to modify its routing topology, you end up with things getting better or worse depending on people, but :
- the isp sometimes has very little leverage on this
- it's difficult to monitor routes like that from end to end, so chances are the ISP doesn't even notice such impacts.

Also, this can be very random from a day to another as the routing protocols used are a mix of link state and path vectors ones. Each next-hop is decided based on a set of rules (avg load, losses, avg latency, buffer queues)

TL,DR;
ISPs and server providers have really less leverage on this than people would think.

ahh love telecommunication networks ;)

thank you for explaining very well what I know but cant name :D

was hoping someone would show up who knew the names of such companies and had the knowledge to give better examples of how everything is very much more complicated then it is when it comes to routing.

so who was phone?

genuine laugh, thanks

This thread is too technical for me to understand, but I am enraged!

VPN stuff in a nutshell:

if you wanna try one, you have 3 options to consider:

1- vpn server in your own country: to see if you can by pass a bad route your isp takes to leave the country for the destination server

2- vpn server in other country then final destination ip: your country exit hop is fine and pings low but you somehow ping horrible to X country overall, maybe you ping much less to a geographical neighbor and your overall ping is less from connecting to X country from its neighbor.

3- vpn server in final destination ip country: say you ping low all the way to germany from finland, 20 ping, but then you see a hop in germany that takes your ping from 20 to 50. you connect to the vpn server from germany and that server does not have the same problem from your original routing. you only jump from 20 to 30 and all is well.

few final examples i wanted to share.

it's sad that it's 2013 and we have to vpn just to get proper latency to servers.

it's sad it's 2013 and we still have to poop

1 poop removal please

ssssssssssswwhop

speed has increased with 1xxx%
but the ping with my 0,256/0,256 mbit back in 2003 was 25% better than it is today.

Thank marketing department, huh.

First off, without a traceroute from the server to your ip, you are only looking at half the picture. Go here and trace to your ip from the frankfurt site. That won't be exact but will give you a good idea. It may be taking an indirect return path. Second, without traceroutes in both directions when your ping was "good" there is nothing to compare to. It's possible that you were using a transit provider at that point and not direct peering from DNA <-> I3D.

1 46 ms 70 ms 49 ms netnod-ix-ge-b-sth-1500.dnaip.fi [194.68.128.152]
2 52 ms 52 ms 52 ms tuk1-tr1.dnaip.fi [62.78.107.16]
3 57 ms 52 ms 62 ms tur1-tr1.dnaip.fi [62.78.107.50]
4 61 ms 52 ms 55 ms rai1-tr2.dnaip.fi [62.78.107.97]
5 52 ms 52 ms 52 ms rai1-tr1.dnaip.fi [62.78.107.42]
6 51 ms 76 ms 51 ms rma1-tr1.dnaip.fi [62.78.107.100]
7 52 ms 52 ms 52 ms rma1-tr2.dnaip.fi [62.78.107.93]
8 53 ms 52 ms 52 ms rma1-er2.dnaip.fi [62.78.108.33]
9 * * * ?
10 * * * ?
11 * * * ?

yep :(

The first hop is your isps (DNA) address on the netnod exchange in Stockholm. This leads me to believe that i3d does not have a backbone node in Stockholm but is using a long haul connection from Frankfurt to the exchange. My European geography is not so great but looking at google maps I'd estimate that to be around a 25ms distance. Someone else could probably correct me if I am wrong. Netnod's website has i3d's connection listed at 1 gig and DNA's listed at 2x10gig. 2 things you can try from here to see if it is just a congestion issue. First, ping www.netnod.se from your house and see if you get good latency to that site. If its also high, the problem could be congestion on your isps exchange links. If that pings well, then it is likely i3d's end. Another thing to test is to run some pings at off hours and see if the latency is better. If it is good at off hours its likely a congestion issue. It's possible that neither link is congested but that the transport long haul of i3ds is taking a very indirect route.

It's unlikely that your isp nor i3d would stop using this path due to it being a direct free peer. Convincing your isp to get a connection into DE-CIX might help if they were willing to carry the traffic there vs Netnod (a possibility but not 100% likely. Isps prefer to dump traffic at closest exit, which would still be Netnod in this case). Other than that it would just be a matter of waiting for i3d to improve their connection from Frankfurt into Netnod.

im pinging 6 to netnod.se. we're doing some research and it seems this router/whatever is an ongoing issue in other games aswell, think we'll contact i3d.net at some point

interesting, thanks.

Tracing the route to IP node dslb-094-219-111-209.pools.arcor-ip.net(94.219.111.209) from 1 to 30 hops

1 <1 ms <1 ms <1 ms xe--11-3-0.edge7.frankfurt1.level3.net [195.16.162.89]
2 1 ms <1 ms <1 ms vlan80.csw3.Frankfurt1.Level3.net [4.69.154.190]
3 <1 ms <1 ms <1 ms ae-83-83.ebr3.Frankfurt1.Level3.net [4.69.163.9]
4 3 ms 3 ms 3 ms ae-46-46.ebr1.Dusseldorf1.Level3.net [4.69.143.169]
5 4 ms 4 ms 15 ms ae-11-3104.car1.Dusseldorf1.Level3.net [4.69.161.154]
6 7 ms 16 ms 6 ms 212.162.17.22
7 7 ms 15 ms 6 ms 92.79.200.145
8 6 ms 16 ms 7 ms 188.111.129.237
9 22 ms 24 ms 12 ms 92.79.214.118
10 6 ms 16 ms 25 ms 92.79.212.50
11 4 ms 14 ms 4 ms 145.254.1.58
12 23 ms 24 ms 24 ms dslb-094-219-111-209.pools.arcor-ip.net [94.219.111.209]

http://incloak.com/
/thread

Servers go online and offline randomly all the time. It's a disgrace.

Examples of routing
https://www.sprint.net/network_maps.php

http://www.esreality.com/post/1932686/i3d-net...h-related/

thats the post i made when my ping started increasing a couple years ago, it got as bad as steady 65ish where i basicly didnt feel like playing at all, but after some time the route changed and it got settled to 38 for quite a long while, until it has now been steadily increasing now again

sucks, but at least the ping you're getting is still good and playable.

Lots of others can't say the same when this kind of shit happens to them :(

The QL servers are extremely ghetto.

Explain this:

traceroute to 91.198.152.137 (91.198.152.137), 30 hops max, 60 byte packets
1 Warszawa-bras3.inetia.pl (195.114.190.155) 6.352 ms 7.655 ms 8.827 ms
2 WarsH002RT53-WarsH002RT22.inetia.pl (83.238.250.158) 10.293 ms 14.257 ms 14.458 ms
3 xe-1-3-0.waw10.ip4.tinet.net (77.67.65.149) 14.454 ms 15.688 ms 17.152 ms
4 xe-3-3-0.fra23.ip4.tinet.net (89.149.182.221) 38.977 ms 40.106 ms 41.338 ms
5 as3549.ip4.tinet.net (213.200.65.150) 214.414 ms 214.411 ms 214.425 ms
6 interactive-3d-b-v.tengigabitethernet7-2.ar5.fra3.gblx.net (159.63.22.134) 53.857 ms 47.645 ms 48.152 ms
7 quakelive-hosted-by.i3d.net (91.198.152.137) 47.659 ms 47.255 ms 44.912 ms

47 ms ICMP ping, should be 50+ in game, but:

TIMENUDGE!! XERP! SMOOTHCLIENTS!! CVAR SORCERY

Maybe ICMP has too low a priority compared to UDP on some router you go through.

spot on

lol, my traceroute shows 47 ping, also pinging server shows 47 ping as well.

But I have 55 in game!

This is normal. There is a "built-in" lag with the game. The server works at 40 fps so your packet can sit waiting there in queue for as much as 25 ms instead of triggering an immediate reaction.

Incorrect.

Yes, server sends you 40 packets per second. But it doesn't matter. What you see on scoreboard is how the server measures your ping, not the other way around.

So server sends packets depending on sv_fps. But how many YOU send depends on combination of your FPS and cl_maxpackets.

In normal circumstances (125 FPS, com_maxpackets 125) you send 125 packets per second, every 8 ms.

So each 25 ms you receive a packet from server. Then you will include a timestamp or whatever in next packet you send. This can happen immediately, or almost full 8 ms later.

The lag between receiving server packet and sending next client packet changes all the time (no common divisor between 25 and 8), but on average it will be 4 ms.

He has difference of 8 (55-47), which can mean 60 FPS (vsync?) or cl_maxpackets 63.


TL;DR: it adds 4 ms to ping.

t0 : server sends gamestate n
t1 = t0+serverToClientTransferTime : client receives gamestate n
t2 = t1+[0,8] : client treats gamestate n and sends response with lastReceivedState = n
t3 = t2+clientToServerTransferTime : server receives a packet with lastReceivedState = n
t4 = t3+[0,25] : server treats this packet and sends a new one with gamestate n+x

So you're telling me that id tech 3 calculates the ping by doing t3-t0 instead of t4-t0? It is the reception time and not the treatment time?

Correct!

Because that's what ping is about - network transfer time :) The added <0;8) ms delay (on maxpackets 125) is necessary evil to keep the data rate low.

Finally not going through tinet.net anymore, now they peer with PLIX :)

traceroute to 91.198.152.137 (91.198.152.137), 30 hops max, 60 byte packets
1 Warszawa-bras3.inetia.pl (195.114.190.155) 7.042 ms 8.735 ms 9.516 ms
2 WarsB010RT06-WarsC001RT04.inetia.pl (81.210.127.174) 20.011 ms 20.263 ms 21.773 ms
3 WarsH002RT22-WarsC001RT06.inetia.pl (83.238.250.5) 18.323 ms 18.766 ms 21.764 ms
4 interactive3d.plix.pl (195.182.218.80) 40.965 ms 41.134 ms 42.411 ms
5 quakelive-hosted-by.i3d.net (91.198.152.137) 43.637 ms 45.103 ms 46.306 ms

Again ICMP ping 45, scoreboard ping? 31 now :)

(Yes, surely a matter of UDP over ICMP priority.)

Ah, I get the other way around for Chicago's server now. I gotta figure out a way to find out where the UDP traffic gets fucked up.

Constantly between 300 and 999 IG yet a traceroute gives that :

1 17 ms 1 ms 1 ms 10.50.72.1
2 1 ms <1 ms 1 ms 10.2.252.77
3 1 ms 1 ms 1 ms
4 2 ms 8 ms 9 ms
5 1 ms 8 ms 8 ms
6 8 ms 14 ms 14 ms imtrl-rq.risq.net [192.77.63.49]
7 5 ms 5 ms 6 ms te0-1-0-1.ccr21.ymq02.atlas.cogentco.com [38.104.154.61]
8 13 ms 17 ms 17 ms te0-0-0-7.mpd22.jfk02.atlas.cogentco.com [154.54.7.57]
9 20 ms 24 ms 21 ms te0-1-0-3.mpd22.dca01.atlas.cogentco.com [154.54.6.5]
10 21 ms 27 ms 21 ms te0-0-0-1.mpd22.iad02.atlas.cogentco.com [154.54.26.118]
11 21 ms 26 ms 25 ms sl-st31-ash-.sprintlink.net [144.232.8.13]
12 28 ms 28 ms 28 ms sl-crs1-dc-0-13-0-0.sprintlink.net [144.232.19.231]
13 31 ms 30 ms 29 ms sl-crs1-rly-0-1-5-0.sprintlink.net [144.232.20.15]
14 31 ms 30 ms 31 ms 144.232.4.184
15 31 ms 31 ms 32 ms sl-crs1-akr-0-3-0-0.sprintlink.net [144.232.18.59]
16 30 ms 31 ms 31 ms 144.232.1.102
17 40 ms 35 ms 34 ms sl-nlaye-467201-0.sprintlink.net [144.223.73.210]
18 * * * Délai d'attente de la demande dépassé.
19 34 ms 34 ms 33 ms as23352.ae6-102.cr2.ord1.us.nlayer.net [69.31.111.3]
20 34 ms 40 ms 34 ms ae12.cr2.ord6.us.scnet.net [204.93.204.83]
21 35 ms 35 ms 37 ms ae4.ar5.ord6.us.scnet.net [204.93.204.65]
22 35 ms 35 ms 35 ms as36352.xe-2-0-2.ar5.ord6.us.scnet.net [50.31.151.170]
23 33 ms 34 ms 34 ms 10ge-1-3-2.gf146.cr2.chi.colocrossing.com [108.174.54.222]
24 34 ms 34 ms 34 ms host.colocrossing.com [199.21.113.50]

so much geek talk

http://www.quakelive.com/forum/showthread.php...-Frankfurt

First i sent mail about changed routes to info@i3d.nl...Waited 3 days for answer. Then sent to noc@i3d.net and noc2@i3d.net. Answer was from stefanideler@i3d.nl and after some messages he applied another route to my provider and it helps. May be this will help somebody. Now ping is stable 32ms from Saint-Petersburg, Russia.

WoW they actually changed it? Faith in humanity is restored!

a bit...as they can :)

someone needs a good route

Well I have documented to the best of my ability over fifty player cases, on the bad NetNode, and send mail to i3d with a good description of that is going wrong.

Now we can only hope they will take it seriously.

If anyone is intrested here is the Mail and attachment

I am getting the impression alot of players from finland and denmark end up using that node so you have a good case going here!

They actually replyed already.

like we talked on irc, the issue here is partly that its the same route as i had before with 38 ms, now just the netnod gives 20 more than before..

confusing answer. this answer means that you already arrive to the destination country at those nodes right? still it doesnt explain why exactly you ping much lower to somewhere in germany, frankfurt when its not that i3d router?

Maybe I am misunderstanding this but it still sounds like you get +20 ping for no good reason to either nl or de i3d.

It's not that confusing, when you look at the path the packets take from I3D back to telenor (in amOKchen's case):

Router: fr0.ix1.frm.as49544.net
Command: traceroute 94.144.1.139

Type Control-c to abort
Sending DNS Query to 213.163.76.185

Type Control-c to abort
Tracing the route to IP node lo0-1.bb03-jd395.ip.telenor.dk(94.144.1.139) from 1 to 30 hops

1 15 ms 7 ms 18 ms 109.200.218.82
2 18 ms 7 ms 17 ms ti9000b300.telenor.net [195.69.145.192]
3 40 ms 38 ms 37 ms ti3004c310-ae2-0.ti.telenor.net [146.172.105.53]
4 52 ms 75 ms 36 ms ti3004b300-ae0-0.ti.telenor.net [146.172.105.42]
5 43 ms 36 ms 38 ms cybercity-2.ti.telenor.net [148.122.9.26]
6 40 ms 49 ms 49 ms lo0-1.bb03-jd395.ip.telenor.dk [94.144.1.139]

You can see that the packets are sent directly to telenor from i3d's side. Telenor on the other hand insists that to get to the server, you should take the route through Netnod's internet exchange (at least in the case of the subnet range appointed to the area where amOKchen resides) which means they go to Stockholm and from there through the i3d's backhaul to Frankfurt which isn't needed since they have a direct connection to the Frankfurt internet exchange. The way I see it is that telenor can fix this or give more insight (since there's no looking glass available) on the issue but will they do anything about it? I doubt it, but it doesn't hurt to ask.

sounds "too much embedded in the system" for anyone at telenor to change?

The basic rule of thumb regarding the "internet" routing is that you can't tell the neighbouring AS (ISP, Server host etc) how to route packets in their AS (you can try to influence it from your side, but it usually has little effect), everything that happens inside an AS is the responsibility of the company that holds the said AS number.

Main obstacle here to make the changes for an ISP is the cost (in addition to design constraints, corporate routing policies etc). ISP-s don't offload traffic to the nearest exit, they offload it to wherever they have the biggest and cheapest pipes to (the side effect of increasing download speeds). Depending on the size and location of the ISP, it's usually the 3 largest internet exchanges in Europe: Frankfurt, Amsterdam and London.

So if they have 40Gbps (for example) to Netnod and they've directed let's say half of Denmark through there, there's very little chance that they would change their routing policy just for a handful of gamers (they can't just reroute 1 IP, the administration overhead it generates is ridiculous).

With that being said, it still doesn't hurt to ask :)

I see, yes thats what I figured for individual ip routing changes here in turkey too :(

so it helps u?

No, he just told us that the two nodes in question is not in Sweden.
But what must be a direct fiber route between stockholm and frankfurt/amsterdam.

We still sit here with all our problems.

Not sure all Danes and Swedes get the problem. I'm on Syd Energi 40/40 and the traceroute is as follows.


Sporer rute til quakelive-hosted-by.i3d.net [91.198.152.137]
over et maksimum af 30 hop:

1 <1 ms <1 ms <1 ms 192.168.1.1
2 <1 ms <1 ms <1 ms 95.154.12.131
3 1 ms 1 ms 1 ms 89.184.134.98
4 1 ms 1 ms 1 ms 87.116.1.161
5 1 ms 1 ms 1 ms 87.116.1.160
6 22 ms 22 ms 18 ms 10ge.de-cix.i3d.net [80.81.194.107]
7 18 ms 18 ms 18 ms quakelive-hosted-by.i3d.net [91.198.152.137]

Sporing fuldført.

Im having the same problem in belgium.
Ping on .de was 24-30 and now it randomly changed to 70-80.

Meh...

From Bulgaria 32ms ping before to de.. after that it went to 45 and now 52-55.. damn

lol euronoobs whining about 50ms pings, sad

normally I would give the same answer because of my situation but they do have a point because they CAN get much better usually :/

said the guy that pings 10 ms to moscow and 40 ms to frankfurt (you can see bellow in comment #94 why selfish h8m3 doesnt want any changes of the current situation).

the europeans whine about unstable 50 ms, because they are used to low pings and stable connections. for the same reason they always whine, when their tournaments are joined by players from far away (like east-russia) and they have to play on "fair", unstable and east-friendly servers with pings 50+ ms and warp.

you would whine aswell, when you would have unstable 50 ms to moscow, while you should have a much better connection than that. and in such a situation the last thing you want to read, is something like comment #78 where you get called a sad noob.

Hi,

I think it goes without saying that this routing you're having there is bullshit :)

Of course at first glance it seems that the node in Stockholm is the one to be blamed, which might in fact be the case.

But there is also another possibility, which has not been addressed here. I'll try to explain the concept to you, since I don't know the correct terminology for it: Some carriers use some kind of tunneling between nodes. For example, my own ISP does it, since my first hop outside my house is displayed being in Frankfurt (which is 650km away). That said, I also observed this kind of tunneling with other carriers. Some in-between-nodes are simply not displayed. I suspect you're in fact routed via Stockholm -> Dusseldorf (or somewhere close to it) -> Amsterdam -> Frankfurt. Maybe even via London. Some carriers simply have fucked-up routing like that.

The high ping to the Stockholm node is obviously wrongly displayed (have seen that before as well), as you appear to ping 50 to Stockholm, but 45 to Frankfurt. For that there may be two causes. First, seeing that the Stockholm node pings higher than the succeeding node in Frankfurt, which is impossible, clearly shows ICMP traffic is prioritized low, so you might as well have a good "real" ping to Stockholm. Second, the Stockholm node displays the ping to Frankfurt / Asmterdam / whatever. I have no conclusive explanation for that, but again, I've seen that before.

Long story short, I think Stockholm node is fine, but you are routed via more hops than you can see.

It's not really tunnelling in the same sense as you might think a vpn is, but close.

Here goes: Networking is based on the OSI model. Simply put it consists of layers on top of each other, starting from the foundation (imagine building a house, foundation first, add walls etc). Traceroutes give you information about layer 3 which basically is IP information but there are two more layers below which play an equal part in the connection. The invisible nodes in case of ISPs are the ones which transport the data between two routers. In case of transit traffic this usually takes the form of a DWDM network (layer 1). The main goal of those nodes is to transfer packets from one side to the other over fiber. Those nodes do not care about the payload inside the packets. Physics do apply though and considering the length of the DWDM path the delay can add up (in the form of how much distance the light signal has to cover in order to reach the other side). What determines the DWDM path? Again, it is most likely the cost. The ISP has to pay some other provider to use their DWDM network to get the connection to wherever they want (internet exchange) if they can't get there with their own equipment.

One more thing, the geo IP locations you see are not really accurate. You can't really tell where the router is physically, looking at the IP address alone, so take the geo IP location information with a grain of salt.

Just my 2 cents.

DWDM is just lower lever transport, but above it ISPs usually use technology called MPLS, which is also a sort of tunneling that can hide internal ISP structure and can produce weird things with traceroute (which are actually most likely reason of big jump in ping at Stockholm hop in OP's traceroute reading instead of some problem with router as he suspects).

Basically, inside ISP network a set of MPLS "label switched paths" (LSPs) is built between edge routers. Then when packet enters the network, router puts a label on it, and then instead of being routed based on destination IP address, packet is forwarded based on its attached label along one of those LSPs.

Traceroute works by limiting the amount of hops that packet is allowed to travel through the network and then recording at which router the packet reached that limit (that router will send an ICMP message back to source indicating that limit is exceeded). Then by sweeping this limit from 1 to some large value, you will "probe" the routers at various distances along packet path, and thus determine through which routers it passes.

Now where MPLS breaks traceroute is in the middle of LSP. Usually routers in the middle don't know any internet routes. So if they produce an ICMP message, they won't know how to send it back to source. Because of this, a trick is used called "ICMP tunneling", which, basically means that instead of sending ICMP message back to source, router attaches the same MPLS label to ICMP packet as was in original IP packet and send it forward towards the end of LSP. Then, when it reaches the end, edge router will have internet routes and will be able to properly send it back. So instead of packet going to a router and then ICMP message going back from it, packet will go to a router, and ICMP message will go further forward until the very edge of ISP, then it will turn around and head back through the whole ISP network and finally return, which takes a lot more time obviously.

Thus, basically, if MPLS and ICMP tunneling is used in ISP network, no matter where the hoplimit-breaking router is in that network, data will always travel full way forward and back through the whole diameter of ISP and you will see same ping value for all routers belonging to that ISP, which is also much higher that actual time it takes for packet to travel normally.

And as you can see in OP's traceroute, he sees basically the same exact ping for Stockholm and Frankfurt which are separated by about 1000km along the path, which is hard to explain without involving behaviour described above.

(Also, for most routers in the ISP network, TTL decrement is usually disabled, so you don't see them in the path at all, that's why those 1000km are seen as if it was just 1 hop, though probably there are several intermediate routers as well)

Yes, MPLS works at layer 2 (or layer 2.5, however you want to call it) and the lower layer transport is just as important delay wise. You don't need multiple routers along the 1000km path. That's one of the reasons why you'd use DWDM in the first place. You have one router in the internet exchange connected to core of the ISP as if there isn't 1000km (or more) in between( and yes, on top of DWDM you can run MPLS or whatever you choose).

The reason why the OP has the exact same ping for Stockholm and Frankfurt, is because the router which you think resides in Stockholm, actually resides in Frankfurt as the e-mail from i3d clearly shows. So the packets from the finnish ISP first terminate in their Netnod router (most likely, no looking glass to verify this), and through the Netnod's GigE switch (http://www.netnod.se/ix-stats/14all_ix-load-p...olm.geb.26) go to i3d, which in that case is a lower layer transport bought by i3d from Stockholm to Frankfurt (DWDM most likely).

So it's really not that hard to explain without ICMP tunneling in the picture.

I got 40ms on DE from Moscow now which is my lowest ping ever there, leave it as it is now!

you live in moscow and got servers in moscow where you should ping about 10 ms. the de servers arent dedicated to you.

I ping 1-2ms to Moscow :-)

NICE.
must be like playing on lan everytime, while being at home.
and in a pretty jealous manner i would like to add: asshole
>:]

My ping changes after every ql update. First i had 45 then 35, 45, 55 and now 60 on .de servers. I have same issue as frs and amokchen with a big hop on traceroute. Tried using iclock vpn with different locations. Best result was ping 45, but in ql it gives massive lagspikes, and is even worse than ping 60.

unfortunately that depends entirely on the load of the vpn server and how well it is mantained.

For example I can get 50 ping to romanian qlive servers with incloak's vpn server. But the servers are so horribly loaded that it is impossible 99% of the time. Fortunately I don't have the same problem with the server I use to lower my ping to the DE servers!

you can try contacting incloak, they have a surprisingly reliable customer support.

I sound like a incloak rep in this thread lol

Raist: "Fortunately I don't have the same problem with the server I use to lower my ping to the DE servers!"

i couldnt believe my eyes as i saw this.
raist always warps like hell on de servers.
he always denies that he is warping (see the lie with exclamation mark above).
all players complain about his warps and he already had been kicked from public games a couple of times due to his lags.
i always thought it is because his turkish stone-age computer cant get over 5 frames per second, turkish internet infrastructure is medieval or his turkish mother has a longer beard than him.

now this monkey reveals that he uses a vpn, while every idiot knows that this causes laggy connections.
and that is not all: he dares to give false advices on connections.
and it is getting even better: he seriously recommends other players to use vpns too.

\bind space "callvote kick Raist; say reason: warping vpn user; wait 1000; say bye bye retard and have fun on servers in your third-world-vicinity"

"all players complain about his warps and he already had been kicked from public games a couple of times due to his lags."

stopped reading there. sorry lying wont get you anywhere, don't really know why you are obsessed with me about this. I suggest you deal with whatever issue (jealousy? me fragging you? heck if i know) you have.

"sorry lying wont get you anywhere" said the monkey, that connects from turkey to germany over a vpn and tries to make people believe he doesnt warp. hahahahaha.
and i remember you getting kicked for warping a couple of times damn well, because i was the one that called the votes. hahahahaha, massive fail raist.

and why should i lie in comment #109 about something everyone can check? just connect to a de server with raist on it to see him warp, players complain about that and players try to kick him.

and why should i be jealous? because you got more hair on your ass than a dog and live in a country where the women look like klingons? hahahahaha, sure.

tl;dr.

i don't warp, get help mate.

end of story.

"tl;dr." - why you reply on a post you didn't read? this doesn't make any sense. how you want to commentate on content that you don't know? massive fail, klingon!

"i don't warp, get help mate." - sorry lying wont get you anywhere, except on youtube:



"end of story." - the end of story is reached when i tell you so, because pets have to obey their owners or they will get castrated.

\bind space "callvote kick Raist; say reason: warping vpn user; wait 1000; say bye bye retard and have fun on servers in your third-world-vicinity"

END OF STORY.

hahahaha

use star trek music for the next one!

still tl;dr btw, not worth my time

i was shocked that comment #120 got plussed by two idiots, eventhough it is full of shit.

"me":
made him ragequit last time i saw him, as i revealed that it is a fakeaccount of the tierslumming statslamer "metapsy". cry more faggot.

"h8m3":
hates me because i revealed that he should stfu in this thread. h8m3 tell me please, since when russians are allowed to judge warp related content?
you connect from moscow to a server in moscow with a ping of 8 ms and you still manage to lose packets on such a short distance. it is on the lagometer in your own videos.
example: http://de.twitch.tv/h8m3/b/308868564
and you want to judge if raist warps on ger servers? hahahahaha.
(by the way, wp.)

learn to judge posts by their content, esr-monkeys.

Stop projecting so hard, mon0.

You're a notorious idiot, rage quitter, and tierslumming fag, not the other way around.

I left a random public server during warmup last time you saw me and it had nothing to do with yourself nor your 'revelations', sorry to disappoint.

"mon0" - ?

"You're a notorious idiot, rage quitter, and tierslumming fag" - lol.

"I left a random public server during warmup" - lol. was one of the funniest ragequits i ever saw.

sorry but raist doesnt warp, and me isnt tierslummer...
i do remember you though, you have some mental issues :/

you need some love, fresh air, and stuff like that
good luck

"sorry but raist doesnt warp" - since when poeple from hungary are allowed to judge warp?

"me isnt tierslummer" - i checked the player ratings before (mine) and after (his and mine) the game and the qlranks history. he was a tier 4 dueler and joined a server where only i was on (bloody duel noob), so server was tier 3 (like me at that time point). i told him to gtfo during warmup coz i already saw it would result in 0-30, but he readys up instead to improve his stats. lacceh, what qualifies you to talk about that?

"i do remember you though" - yeah, i am the guy at the other end of the projectiles that keep fragging you.

"you have some mental issues" - said the guy, that obeys danstar. HAHAHAHAHA.

i ping 80+ to russian server on ql and 20 at q3defrag :o

finland > us > china > uk > ger > aus > russia :DAAA

davai davai

Same shit here.

1 <1 ms <1 ms <1 ms ip1.grubbehus1.ac.se [77.110.0.1]
2 2 ms 1 ms 1 ms metro1g5.umea.net.riksnet.se [85.11.2.237]
3 <1 ms <1 ms <1 ms city1r.umea.net.riksnet.se [85.11.2.81]
4 8 ms 8 ms 8 ms stockholm1rc.net.riksnet.se [85.11.0.1]
5 9 ms 9 ms 9 ms stockholm2rc.net.riksnet.se [85.11.0.2]
6 73 ms 74 ms 74 ms netnod-ix-ge-b-sth-1500.i3d.net [194.68.128.206]

7 67 ms 67 ms 68 ms quakelive-hosted-by.i3d.net [91.198.152.121]

same problem different location.

i am routing with fastpath from berlin germany.
normal ping to de/nl was about 30 ms.
non-normal ping to de/nl is about 60 ms.

in the past i experienced the the routing to fuck up 3 times (after update 12, 2 months in fall 2012 and right now).

in my case: tracerts have shown the route doesnt take any non-normal hops in other countries, the dramatic ping increase is located at the last hops that are already arround frankfurt (aka the "last mile").

and dont listen to raist: when other non-ql-servers in frankfurt/amsterdam still ping 30 ms, while ql servers there ping 60 ms, the problem is caused by the ql server hoster's "last mile" and not your isp.

so i live in berlin germany and currently ping 60 ms to ql frankfurt/amsterdam and 40 ms to ql stockholm/milan/maidenhead.
fu i3d.net.

yep, this is probably my #1 issue with ql. ping can change by 30-35 overnight and you're basicly playing a different game after that. and the answer is "its working fine" lol.

and yep, i've had the same exact route for the last 6-8 months or whatever and i've had 38 steady ping, now for some reason this router has 25ms higher ping... amokchen is on a holiday or something but after hes back we'll contact i3d again

4 7 ms 6 ms 6 ms rai1-tr1.dnaip.fi [62.78.107.101]
5 6 ms 6 ms 6 ms rai1-tr2.dnaip.fi [62.78.107.43]
6 6 ms 6 ms 6 ms tur1-tr1.dnaip.fi [62.78.107.96]
7 6 ms 6 ms 7 ms tuk1-tr1.dnaip.fi [62.78.107.51]
8 6 ms 6 ms 6 ms tuk2-tr1.dnaip.fi [62.78.107.17]
9 46 ms 46 ms 46 ms netnod-ix-ge-b-sth-1500.i3d.net [194.68.128.206]
10 45 ms 46 ms 46 ms quakelive-hosted-by.i3d.net [91.198.152.121]

today it seems to be 10 lower than last week

contacting i3d is prolly the best way to go, because your own isp cant make any drastic changes at the "last mile" of the routing, it mainly depends on the server hoster.
i hope there will be a solution that will lower and stabilize pings for all players (even h8m3). (my main problem aint the 60 ms instead of 30 ms, it is the stability at the ping increasing hops.)

I had +30 ping to de/nl for a 1-2 months or maybe even more, since I had a break from Quake for 4 months or so.

Since a week I have my normal ping back - 35de/40nl. So something is definitely fucked up. Now it's in my favour, next month could be in yours... ;]

i hope i3d.net's routing-lottery stops in a manner that makes the majority low-pinged and stable.

3 6 ms 6 ms 6 ms rma1-tr1.dnaip.fi [62.78.107.92]
4 8 ms 9 ms 12 ms rai1-tr1.dnaip.fi [62.78.107.101]
5 6 ms 6 ms 6 ms rai1-tr2.dnaip.fi [62.78.107.43]
6 6 ms 7 ms 7 ms tur1-tr1.dnaip.fi [62.78.107.96]
7 6 ms 6 ms 6 ms tuk1-tr1.dnaip.fi [62.78.107.51]
8 7 ms 7 ms 6 ms tuk2-tr1.dnaip.fi [62.78.107.17]
9 37 ms 42 ms 38 ms netnod-ix-ge-b-sth-1500.i3d.net [194.68.128.206]
10 37 ms 37 ms 37 ms quakelive-hosted-by.i3d.net [91.198.152.137]

WOO time to play some quake again! :D

3 2 ms 2 ms 2 ms 193.166.149.1
4 4 ms 4 ms 4 ms csc6-x11200-utu6.funet.fi [193.166.255.97]
5 11 ms 11 ms 11 ms se-fre.nordu.net [109.105.102.65]
6 52 ms 49 ms 49 ms netnod-ix-ge-b-sth-1500.i3d.net [194.68.128.206]
7 46 ms 46 ms 46 ms quakelive-hosted-by.i3d.net [91.198.152.121]

trying from uni, same issue here. 50 ms isnt that awful but nothing really explains the added ping, i've had 29 ping to this server before (back in the good old days) so the issue isnt distance or germany in general, i can get ~30 ping to .de np

-frs

this looks fun, what ip do i tracert and how do i copy all the text from the cmd window?

the ip you see when you join a server


to copy all the text just right click and select all

Thanks!

Spårar väg till quakelive-hosted-by.i3d.net [91.198.152.211]
över högst 30 hopp:

1 1 ms 1 ms 1 ms mah shitty old router c:
2 8 ms 9 ms 7 ms c213-100-116-1.cust.tele2.se [213.100.116.1]
3 11 ms 19 ms 12 ms lse8-tvds-1.vlanif7.tele2.net [130.244.9.189]
4 14 ms 12 ms 30 ms ume1-pe-2.tengigabiteth9-1s465.tele2.net [130.244.35.73]
5 12 ms 13 ms 13 ms ume1-ncore-3.tengige0-0-0-1.tele2.net [130.244.71.180]
6 19 ms 17 ms 19 ms sva-core-3.pos8-0-0.tele2.net [130.244.71.184]
7 20 ms 19 ms 19 ms sva-core-1.tengige0-1-1-3.tele2.net [130.244.71.221]
8 24 ms 24 ms 24 ms kst5-core-1.tengige0-2-0-0.tele2.net [130.244.71.200]
9 25 ms 24 ms 24 ms kst5-peer-1.ae0-unit0.tele2.net [130.244.205.125]
10 24 ms 25 ms 25 ms avk6-peer-1.ae1-unit0.tele2.net [130.244.39.11]
11 24 ms 25 ms 25 ms te4-3-10G.ar2.ARN3.gblx.net [64.208.110.149]
12 45 ms 45 ms 46 ms interactive-3d-b-v.tengigabitethernet7-2.ar5.fra3.gblx.net [159.63.22.134]
13 46 ms 46 ms 46 ms quakelive-hosted-by.i3d.net [91.198.152.211]

Spårning utförd.

Fine I guess :@

easier to type cl_currentserveraddress, instead of having a few seconds to memorize an ip

"fortunately" QL freezes for like half a minute upon connecting now anyway :@

haha true, that's a "feature"!

Ok so it's not just me... Keep having these delays from 2 PC since last update I think :s

i haven't this problem anymore since last update, do you?

nope all good now!

thanks.

1 <1 ms <1 ms <1 ms Broadcom.amis.lan [192.168.1.254]
2 1 ms <1 ms 1 ms asr-lj.amis.net [212.18.32.174]
3 <1 ms <1 ms <1 ms mx-lj1-te-0-3-0.amis.net [212.18.44.161]
4 8 ms 8 ms 10 ms mx-vi1-te-0-0-0.amis.net [212.18.44.142]
5 22 ms 23 ms 22 ms 10ge.de-cix.i3d.net [80.81.194.107]
6 22 ms 22 ms 22 ms quakelive-hosted-by.i3d.net [91.198.152.137]


It's ok from Slovenia. We had some issues many months ago. Nothing as of late tho.

OK from today I suddenly got ping from 45-50 on DE to 120 or more.

Tracing route to 91.198.152.137 over a maximum of 30 hops

1 <1 ms <1 ms <1 ms 192.168.1.1
2 * * * Request timed out.
3 12 ms 11 ms 11 ms 81.93.65.189
4 * * * Request timed out.
5 83 ms 83 ms 82 ms 79.101.106.125
6 * * * Request timed out.
7 110 ms 118 ms 102 ms 212.200.5.113
8 * * * Request timed out.
9 108 ms 113 ms 110 ms 80.81.194.107
10 117 ms 116 ms 116 ms 91.198.152.137

Trace complete.

But not on all QL servers as you can see:



Several DE servers, FR, etc.

Fucked up completely :/