I'm getting 100 choke and a 2000 ping during combat on the MeatHook, and I'm pretty sure the problem isn't at my end.
:x :x :x :x :x :x
I've tried 3 different ISP's, and I've nto had any similar problems on other CS servers.
Anyone know whats going on? I seem to be the only person affected :x
www.silicon.com/a56074 (http://www.silicon.com/a56074)
It's not you Isp. I'm getting it on BY :(
Well I know it's not the server.......must be summint in da middle...
Contractors managed to rip out a major duct route in Manchester yesterday. A large number of high bandwidth bearers were cut in the process resulting in major re-routing of traffic.
definatly not the isp tut as i have the same 1
I'm told all the Manchester stuff was meant to be fixed yesterday, but I'm still getting problems :(
So lets look at the facts:
1) Sheepy and Tutonic have the same ISP
2) All physical problems concerning Manchester, contractors and the like have been recified
3) Sheepy has not reported having the same problems as Tutonic
4) I get no porblems whatsoever with the server from my ISP using a level of "rate 6000"
5) Although some people have problems when on meat hook, most don't so the server appears to be in fine shape
I can only conclude that Tutonics pc is shagged in some manner :?
Sorry mate, have you done the dreaded CS reinstall from scratch yet?
We're not using the same ISP... although I have tried several different ones....
And the problem just suddenly started, I didn't change anything, so I'm not sure a re-install will help much :(
My guess would be network problems then, I don't normally say that too quickly as I work in networking but to give an example... When playing on a couple of US server I frequent, I get a very good response and smooth play, no loss and little choke, using my current config. These servers are powerful specs like Baldric. Occasionally the choke goes up to 100. At the same time their web page takes ages to load up (separate servers obviously) so I would expect that one hop on your route to the server is playing up either taking the re-routed traffic from an unreliable link elsewhere or is running with intermittent errors.
I did quite a bit of config tweaking and think I've got it ok for both UK low ping servers and US 100ms type ping servers so it means if the UK connection deteriorates then my client can handle it as I am setup to compromise between low and higher ping server connections.
Unfortunately, if data is being lost then you get a little lag and if packets are being delayed getting to your from the server you get that horrible 15 second lag and high levels of choke.
Only solution is change your ISP but BT own so much of the backbone in the UK that if it is a BT link that is killing you it possibly won't matter if you change. I'm with eclipse.co.uk and they have a couple of OC3s into the internet from the BT DSL network and this is for Eclipse customers only, no BT traffic uses this bandwidth (I think).
Send over you're config or post it in here and let me have a look at it.
Unfortunately I've gone through all the checks myself once or twice and found not obvious problems with my system but still had poor performance. So in the end a complete reinstall was the last resort and it tends to work quite well.
Sorry for the big post.
I vote for networking too. Can we post traceroutes to homer to see if there are common threads in the routing (there obviously will be towards the end) but it may allow us to isolate the segment that is causing problems.
TraceRoute - Tuesday, October 29, 2002 14:44:22
Generated by CyberKit Version 2.5
Copyright © 1996-2000 by Luc Neijens
Address: homer.deadmen.co.uk
Number of Packets: 1
Packet size: 1000
Trace Hop 1 through 30
Timeout: 5
Resolve Addresses: No
Don't Fragment: No
#1 (10.125.64.1): TTL Exceeded, ttl=255, 50 ms
#2 (62.31.64.33): TTL Exceeded, ttl=254, 20 ms
#3 (194.117.136.10): TTL Exceeded, ttl=248, 30 ms
#4 (194.117.136.6): TTL Exceeded, ttl=249, 20 ms
#5 (194.117.154.9): TTL Exceeded, ttl=251, 41 ms
#6 (195.66.224.47): TTL Exceeded, ttl=249, 20 ms
#7 (195.157.8.2): TTL Exceeded, ttl=250, 30 ms
#8 (62.169.128.74): TTL Exceeded, ttl=248, 30 ms
#9 (80.71.0.50): TTL Exceeded, ttl=247, 60 ms
#10 (80.71.4.6): TTL Exceeded, ttl=246, 110 ms
#11 (80.71.4.10): TTL Exceeded, ttl=245, 70 ms
#12 (193.110.88.20): Echo Reply, ttl=244, 50 ms
Statistics: Out 12, in 12, loss 0%, times (min/avg/max) 20/44/110 ms
The three 80.x.x.x IPs regularly give high times. They are in the manchester area! (80.71.0.50, 80.71.4.6 and 80.71.4.10) However, high ping times may not point to the culprit as high ping on a lightly loaded segment is not an issue. Can we have a few more traceroutes please?
What the HELL was that about? :(
Looks like a fun thing to do but how do you do it? :?
You can run a trace to a remote server from within a dos window or (as I did) use a bespoke program called Cyberkit (which is free). There are also many other programs that can do this. What it shows is a route from your PC to the destination and indicates a speed of response of each node (router) it passes through. In the above example hop #10 to 80.71.4.6 shows a response time much higher than you would expect which might (and i stress might) indicate a slow node.
From DOS the command you could use is "tracert homer.deadmen.co.uk" and you can then copy the contents of the output to the clipboard. In short, it is part of a standard diagnostic when you begin to trouble shoot a network problem - it doesn't give all the answers but it is a start :-)
QuoteThe three 80.x.x.x IPs regularly give high times. They are in the manchester area! (80.71.0.50, 80.71.4.6 and 80.71.4.10) However, high ping times may not point to the culprit as high ping on a lightly loaded segment is not an issue.
Those 80.71.X.X IPs are one of our providers. Soon I hope to move the server's traffic over to a network with less latency. Whats good for websites is not neccesarily good for game servers.... damn em :-)
Over these last few days things have been a little bumpy, however I've been assured things will improve.
Sounds good to me :-)
Cheers
Tracing route to homer.deadmen.co.uk [193.110.88.20]
over a maximum of 30 hops:
1 159 ms 158 ms 143 ms 192.168.255.1
2 146 ms 164 ms 167 ms 192.168.255.2
3 152 ms 158 ms 167 ms interconnect6-l0.ilford.fixed.bt.net [194.74.77.
230]
4 150 ms 143 ms 167 ms core1-gig10-1.ilford.ukcore.bt.net [194.74.16.11
0]
5 154 ms 167 ms 167 ms core1-pos8-0.telehouse.ukcore.bt.net [194.74.65.
117]
6 153 ms 143 ms 158 ms linx-gw1.uk.netscalibur.net [195.66.224.47]
7 165 ms 167 ms 164 ms gige2-0.br2.th.rtr.uk.netscalibur.net [195.157.8
.2]
8 161 ms 176 ms 158 ms lon1-gw3.uk.boyannet.co.uk [62.169.128.74]
9 177 ms 179 ms 177 ms lon2a.core.rtr.caladan.net.uk [80.71.0.50]
10 198 ms 179 ms 191 ms man1b.core.rtr.caladan.net.uk [80.71.4.6]
11 186 ms 194 ms 194 ms cwgway1.crosswired.co.uk [80.71.4.10]
12 183 ms 173 ms 182 ms homer.deadmen.co.uk [193.110.88.20]
Trace complete.
Looks ok to me...
So far I've tried three different dial-ups, Vispa, Claranet, and Jolt, all have the same problems :(
i seems to get most delay on the crosswire server. which i guess is homer's ISP's server :(
Hop----IP Address ---- Node Name ---- ms
0 ---- 192.168.1.51 ---- games.hometest ----
1 ---- 10.13.0.1 ---- - ---- 9
2 ----195.188.52.17 ----- ---- 9
3 ---- 194.117.136.29 ---- har-bas-pos.telewest.net ---- 13
4 ---- 194.117.136.26 ----tele2-har-pos.telewest.net ---- 12
5 ---- 194.117.154.9 ---- ge51-linx-gw1.cableinet.net ---- 16
6 ----195.66.224.47 ---- linx-gw1.uk.netscalibur.net ---- 12
7 ---- 195.157.8.2 ---- gige2-0.br2.th.rtr.uk.netscalibur.net ----11
8 ----62.169.128.74 ---- lon1-gw3.uk.boyannet.co.uk ---- 11
9 ---- 80.71.0.50 ---- lon2a.core.rtr.caladan.net.uk ---- 11
10 ----80.71.4.6 ---- man1b.core.rtr.caladan.net.uk ---- 47
11 80.71.4.10 cwgway1.crosswired.co.uk 87
12 193.110.88.20 homer.deadmen.co.uk 19
Roundtrip time to 193.110.88.20, average = 19ms, min = 19ms, max = 34ms -- 29-Oct-02 17:47:51
OK here's mine. And I must say Stryker, it looks like the problem is between the crosswired router and the server, so a bit of LAN congestion maybe? I take it the server is co-located? Possibly with something that is getting well used?
C:Documents and SettingsAdministrator>tracert homer.deadmen.co.uk
Tracing route to homer.deadmen.co.uk [193.110.88.20]
over a maximum of 30 hops:
1 2 ms 1 ms 1 ms SpeedTouch.lan [10.0.0.138]
2 19 ms 19 ms 19 ms lo1-bba1.th.eclipse.net.uk [212.104.130.141]
3 19 ms 21 ms 20 ms 212.104.128.101
4 22 ms 21 ms 21 ms fe4-core1.th.eclipse.net.uk [212.104.128.225]
5 25 ms 23 ms 20 ms linx1.lon1.as8553.net [195.66.224.126]
6 22 ms 21 ms 21 ms e2-0.lon6.as8553.net [195.10.255.194]
7 90 ms 52 ms 23 ms s6-1.bris1.as8553.net [195.10.255.134]
8 25 ms 25 ms 24 ms fa0-0-2.bris2.as8553.net [195.10.254.82]
9 29 ms 28 ms 29 ms s6-0.man1.as8553.net [195.10.255.14]
10 30 ms 30 ms 30 ms cwgtway2.crosswired.co.uk [195.10.230.11]
11 368 ms 172 ms 179 ms homer.deadmen.co.uk [193.110.88.20]
Trace complete.
Bizarre!
In fact sorry I appologise, its a dedicated server cause the clan just raised funds for it.....
But there is something going on the the crosswired lan casueing serious increases in response times. I was in MH a minute ago and I had 200ms when I usually get 50ms.
Here's mine, but i like to be diffrent so i use my UO trace utility.
(http://www.nsr.deadmen.co.uk/traceroute.JPG)[/img]
Been trying to get a bit of lag to show so that I can traceroute with lag. I have noticed that sometimes there is a slow response to the forums (is the forum colocated with meathook?). Anyway, when I get the "treacle effect" the traceroute shows:
TraceRoute - Tuesday, October 29, 2002 20:21:45
Generated by CyberKit Version 2.5
Copyright © 1996-2000 by Luc Neijens
Address: homer.deadmen.co.uk
Number of Packets: 1
Packet size: 1000
Trace Hop 1 through 30
Timeout: 5
Resolve Addresses: No
Don't Fragment: No
#1 (10.125.64.1): TTL Exceeded, ttl=255, 20 ms
#2 (62.31.64.33): TTL Exceeded, ttl=254, 20 ms
#3 (194.117.136.10): TTL Exceeded, ttl=248, 70 ms
#4 (194.117.136.6): TTL Exceeded, ttl=249, 30 ms
#5 (194.117.154.9): TTL Exceeded, ttl=251, 50 ms
#6 (195.66.224.47): TTL Exceeded, ttl=249, 50 ms
#7 (195.157.8.2): TTL Exceeded, ttl=250, 30 ms
#8 (62.169.128.74): TTL Exceeded, ttl=57, 70 ms
#9 (80.71.0.50): TTL Exceeded, ttl=247, 30 ms
#10 No response (): , ttl=, ms
#11 (80.71.4.10): TTL Exceeded, ttl=245, 91 ms
#12 (193.110.88.20): Echo Reply, ttl=244, 100 ms
Statistics: Out 12, in 11, loss 8%, times (min/avg/max) 20/51/100 ms
... and our culprit is....(roll of drums)...
80.71.4.6 (Hop 10 for me).
HTH
QuoteQuoteThe three 80.x.x.x IPs regularly give high times. They are in the manchester area! (80.71.0.50, 80.71.4.6 and 80.71.4.10) However, high ping times may not point to the culprit as high ping on a lightly loaded segment is not an issue.
Those 80.71.X.X IPs are one of our providers. Soon I hope to move the server's traffic over to a network with less latency. Whats good for websites is not neccesarily good for game servers.... damn em :-)
Over these last few days things have been a little bumpy, however I've been assured things will improve.[/b]
I missed the part.
So. I gather from all the foregoing extracts from nightmare maths books, that punching my monitor will have a minimal effect on lag. Am I right?
tongue
QuoteSo. I gather from all the foregoing extracts from nightmare maths books, that punching my monitor will have a minimal effect on lag. Am I right?
tongue
Yes - unless you use a stiking force of no less than 15 ftlbs per cm² in the upper right quadrant, of course...
8)
Whoah! Top tip! Thanks I'll try it.
(Thinks: Unless he's trying to use some of his faggoty clown powers on me.)
The clan server is hanging off the same Cisco Cat as the routers. The catalyst switch is hardly being touched by crosswired's paultry traffic.... so why the feckin delay I wonder.....
Of course at this time caladan traffic is a bit hit and miss with some packet loss.... they claim to be working on it, also as I said, soon I hope to move black-adder's traffic over to another supplier.
hmmm interesting... any theories albert?
I'm getting some treacle in BYs network too so I've logged a call for that as well. Should be fixed in a year or two :-(
Stryker, do you need more traceroutes or are you sick of 'em? I know everyone else is :-)
it would be REALLY handy if you could all get a copy of "ping plotter"... its a SUPERB tool for diag of networks.
http://www.pingplotter.com/downloads/pngplt_2.exe (http://www.pingplotter.com/downloads/pngplt_2.exe)
Then do a trace at say any time of your choice and then during a game if your not playing or if your pc is up to it.
set the ping frequency to 1 ping every 2.5 seconds and then run it for at least 15 mins. Then after that under the file menu you can choose to save the data... then mail it to me or the forum.
stryker@deadmen.co.uk
nice screen shot
(http://www.pingplotter.com/nessoftgraph.gif)
I notice that Tutonic, Armitage and I reach Homer on one route (80.x.x.x) while Albert and JB reach Homer through another (195.x.x.x).
Albert and JB show high latency at Homer whereas those coming in through 80.x.x.x seem to show high latency before Homer.
I get huge lag spikes as does Tut. What about you other guys?
Weird!!! :o|:
thanks for spotting that, I've looked at these kinda traceroutes so much the figures all blur into one....
those coming via the 195.. (AS8553.NET) are the fourtunate ones... see how the average response time is so much lower throughout the route... thats the diff a decent transit provider makes...
problem is black-adder does not seem to like talking to our second router (cwgtway2).... weird.
Maybe if our various customers go crazy and pay actually their bills we will be able to afford 3com nics for the routers :-)
QuoteI notice that Tutonic, Armitage and I reach Homer on one route (80.x.x.x) while Albert and JB reach Homer through another (195.x.x.x). Albert and JB show high latency at Homer whereas those coming in through 80.x.x.x seem to show high latency before Homer.
I get huge lag spikes as does Tut. What about you other guys?
I can get loss/choke and lag spikes. And unfortunately it seems that whichever connection I dial up (I have a few options) they ALL route via 80.xx.xx.xx etc. Grrrrr!
TL. 8)
QuoteMaybe if our various customers go crazy and pay actually their bills :-)
It'll never happen :-)
Quoteit would be REALLY handy if you could all get a copy of "ping plotter"... its a SUPERB tool for diag of networks.
http://www.pingplotter.com/downloads/pngplt_2.exe (http://www.pingplotter.com/downloads/pngplt_2.exe)
stryker@deadmen.co.uk
What a cracking toy! How many Gb of data do you want :-)
QuoteQuoteMaybe if our various customers go crazy and pay actually their bills :-)
It'll never happen :-)[/b]
Never underestimate the power of Jo...
:twisted:
I'll run Ping Plotter in the background tonight to record data during the game and see if we can correlate lag spikes with the data.
good idea BB. If the rest of the grunts do it too with enough graphs from enough people covering the same period we may be able to build up a better picture of whats going on.
For comparison, here is graph of it "quiet" :-)
(http://www.mtuspeed.pwp.blueyonder.co.uk/Trace.gif)
You can plainly see packet loss in BY (I have logged this with BY) and Crosswired areas. The red vertical bits are where I, and others, experience lag spikes. It seems to repeat on a 1 minute cycle although sometimes it is so transient it doesn't register. As you can see though, some of the spikes are not so transient and last for anything up to 10 seconds.
This, of course, will be my reason for being killed for the next 6 months so get used to it :-)
Thought I had a screenie of my graph. Thought wrong. I'll be beck
(http://www.finnglish.co.uk/graph.gif)
Let's see now. Yep!
I hope this is helpful, couldn't make head nor tail of it myself :( :? :oops:
God! What a crap screenie! How'd you get yours so clear BB?
It looks like you're fscked before you start, Doorman. PL begins as soon as Blueblunder get their incompetent little fingers on your packets... :(
wow, BB, I get those few second pauses every min or two.... and yes I'm on blueblunder too !!!!
it must be a national thing.
I've also noticed that they seem to cause failed webpages sometimes too. You get page not available but if you hit refresh it appears instantly. At least it is not only BY that have probs. It seems to appear in other areas too looking at the submitted graphs.
QuoteGod! What a crap screenie! How'd you get yours so clear BB?
Yours seems to be stretched somewhat (500x275 pixels compared to mine at 448x391 pixels).
I screengrabbed the window, cut out the bits I didn't want, then saved it as a GIF file - simple as that.
I used the progams 'save as image' then optimsed it using PS 'save for web' and save it as a gif. Still, the point is does it tell you anything? I've been watching various IPs at 2.5 sec. intervals and a very interesting thing it is to. :roll:
QuoteIt looks like you're fscked before you start, Doorman. PL begins as soon as Blueblunder get their incompetent little fingers on your packets... :(
Ker-rist! What am I to do? What am I to do? (runs around in tiny circles clutching at what is left of his hair) Am I wasting my money on the 1mb service? Should I move next door to you? :o|:
The third column (%PL) tells me that 10 percent of the information you are sending/receiving is being lost/delayed/discarded starting at the third hop on the route to homer.
The bottom graph tells me (by looking at the two large red bands) that you had two periods of intense lag.
The round trip information (just to right of middle of screenshot) tells me that your ping times are ok (quite good in fact)
The haziness of the screen shot tells me that the application really screws up screenshots :-)
To take a screenshot of your whole computer screen press the "PrintScreen" key. To take a screenshot of just the active window (which is all you usually need) press ALT and Printscreen together. Either of these actions will put an image into the clipboard which you can then paste into the image editing application of your choice :-)
Quote
To take a screenshot of your whole computer screen press the "PrintScreen" key. To take a screenshot of just the active window (which is all you usually need) press ALT and Printscreen together. Either of these actions will put an image into the clipboard which you can then paste into the image editing application of your choice :-)
Is that the key top right of my keyboard that says 'Print Scrn'? :D
(http://www.finnglish.co.uk/graph.gif)
OK. What does that one say?
If you watch this thing long enough you see every combination of ups and downs.. I've even seen a practically straight line! :smash:
Nice connection, no major probs. Looking at the bottom graphs there were a few occasions where one of the connections had high response times but that is unavoidable and pretty usual. All in all, a good connection for gaming :-)
Awful for me at the moment. Can't even get these forums to load???
I'm running pingplotter. Problems appear to be at:
man1b.core.rtr.caladan.net.uk
Me too i have been trying to get onto these forums for the past half hour and my ping on Homer is fluctuating between 70 and 450..
me too :(
Was gonna play tonight as well.
Two points that are bad...
| Hop | %Loss | IP Address | Node Name | Location | Tzone | ms | Graph
| 1 | | 192.168.7.1 | my.router | ... | | 0 | x | (private use) |
| 2 | | 212.38.173.2 | - | ?(United Kingdom) | * | 26 | x-- | Iomart Ltd |
| 3 | | 212.38.173.2 | - | ?(United Kingdom) | * | 45 | -x------ | Iomart Ltd |
| 4 | | 212.38.173.4 | - | ?(United Kingdom) | * | 29 | x---- | Iomart Ltd |
| 5 | | 212.67.120.90 | - | ?(United Kingdom) | * | 24 | x-- | OneTel.UK ISP Services |
| 6 | | 212.67.120.82 | - | ?(United Kingdom) | * | 20 | x | OneTel.UK ISP Services |
| 7 | | 212.67.120.61 | - | ?(United Kingdom) | * | 36 | -x---- | OneTel.UK ISP Services |
| 8 | | 213.161.69.42 | 90.ge1-0.er1a.lhr4.uk.mfnx.net | ?(United Kingdom) | * | 21 | x- | AboveNet UK |
| 9 | | 208.184.231.129 | so-3-3-0.cr2.lhr4.uk.mfnx.net | ?San Jose, CA 95113 | | 29 | x-- | Abovenet Communications, Inc ABOVENET-6 |
| 10 | | 208.184.231.169 | pos7-0.mpr2.lhr1.uk.mfnx.net | ?San Jose, CA 95113 | | 38 | -x--- | Abovenet Communications, Inc ABOVENET-6 |
| 11 | | 208.185.188.218 | - | ?San Jose, CA 95113 | | 43 | -x----- | Abovenet Communications, Inc ABOVENET-6 |
| 12 | | 193.111.29.22 | fa2-0.lon5.as8553.net | ?(United Kingdom) | * | 51 | -x------ | Bulk Bandwidth UK |
| 13 | | 195.10.255.194 | e2-0.lon6.as8553.net | ?(United Kingdom) | * | 59 | -x-------- | Avensys Networks |
| 14 | | 195.10.255.134 | s6-1.bris1.as8553.net | ?(United Kingdom) | * | 45 | -x----- | Avensys Networks |
| 15 | | 195.10.254.82 | fa0-0-2.bris2.as8553.net | ?(United Kingdom) | * | 129 | ----x-- | UK-GEMSOFT-971114 |
| 16 | | 195.10.255.14 | s6-0.man1.as8553.net | ?(United Kingdom) | * | 36 | -x-- | Avensys Networks |
| 17 | | 195.10.230.11 | cwgtway2.crosswired.co.uk | 53.30n, 2.15w | | 29 | x- | UK-GEMSOFT-971114 |
| 18 | | 193.110.88.20 | homer.deadmen.co.uk | - | | 115 | --x-----