Tweakage help

[Da3dlaus]

Heres the problem: im playing a game that recently switched to a new, "more powerful" server. However, everything is running more slowly, and due to a tool that runs in the background constantly, everything glitches even when there is no lag. This has made it nearly impossible to play against others, and has reduced me from a formidable opponent to a drooling, bumbling n00b. I would love to use the tools available from this site, but all this technical jargon is chinese to me, and my dad wont do it, even though he knows what hes doing (i was 3 times better than him before the server change, and now that im down, he wont let/help me get back up).

In short, i need some serious help in tweaking my connection. Ive done several tests, including one from this site.

« SpeedGuide.net Speed Test Results »
4323 kbps down (~4.32 Mbps, 528 KB/s)↓
175 kbps up (~0.17 Mbps, 21 KB/s)↑
5120 KB downloaded in 9.703 seconds
1024 KB uploaded in 48.031 seconds
Tested on: 2006-10-07 09:25 EDT

Tested from: http://www.digitalmafia.in
Test Validation ID: http://www.speedguide.net/speedtest/...id=F41RXGNUPD2

I have already downloaded and installed the newest version of the optimizer program; now i simply need help using it. If you need more info from me, ill get it up here as fast as i can.

[trogers]

We need the answers to the first 4 questions of this link:

http://forums.speedguide.net/showthread.php?t=73799

Telling us your location (country) will also help.

[Da3dlaus]

1) Cable

2) Cable One

3) TCP options string = 020405b40103030201010402
MTU = 1500
MTU is fully optimized for broadband.
MSS = 1460
Maximum useful data in each packet = 1460, which equals MSS.
Default TCP Receive Window (RWIN) = 256960
RWIN Scaling (RFC1323) = 2 bits (scale factor of 4)
Unscaled TCP Receive Window = 64240

RWIN is a multiple of MSS
Other RWIN values that might work well with your current MTU/MSS:
513920 (MSS x 44 * scale factor of 8)
256960 (MSS x 44 * scale factor of 4) <-- current value
128480 (MSS x 44 * scale factor of 2)
64240 (MSS x 44)
bandwidth * delay product (Note this is not a speed test):

Your TCP Window limits you to: 10278.4 kbps (1284.8 KBytes/s) @ 200ms
Your TCP Window limits you to: 4111.36 kbps (513.92 KBytes/s) @ 500ms
MTU Discovery (RFC1191) = ON
Time to live left = 53 hops
TTL value is ok.
Timestamps (RFC1323) = OFF
Selective Acknowledgements (RFC2018) = ON
IP type of service field (RFC1349) = 00000000 (0)

4) Windows XP Media Center Edition (no service package)

5) Sherman, Texas USA

[trogers]

Do this network diagnostic test and post the results under the 'Statistics' button:

http://nitro.ucsc.edu/

[Da3dlaus]

sorry this took so long,,,my dad was hogging the PC last night.

WEB100 Enabled Statistics:
Checking for Middleboxes . . . . . . . . . . . . . . . . . . Done
running 10s outbound test (client to server) . . . . . 383.06Kb/s
running 10s inbound test (server to client) . . . . . . 4.15Mb/s

------ Client System Details ------
OS data: Name = Windows XP, Architecture = x86, Version = 5.1
Java data: Vendor = Sun Microsystems Inc., Version = 1.5.0_05

------ Web100 Detailed Analysis ------
Cable modem/DSL/T1 link found.
Link set to Half Duplex mode
No network congestion discovered.
Good network cable(s) found
Normal duplex operation found.

Web100 reports the Round trip time = 440.81 msec; the Packet size = 1460 Bytes; and
No packet loss - but packets arrived out-of-order 1.52% of the time
This connection is receiver limited 76.22% of the time.
This connection is network limited 23.76% of the time.

Web100 reports TCP negotiated the optional Performance Settings to:
RFC 2018 Selective Acknowledgment: ON
RFC 896 Nagle Algorithm: ON
RFC 3168 Explicit Congestion Notification: OFF
RFC 1323 Time Stamping: OFF
RFC 1323 Window Scaling: ON
Packet size is preserved End-to-End
Server IP addresses are preserved End-to-End
Client IP addresses are preserved End-to-End

[trogers]

Cable modem/DSL/T1 link found.
Link set to Half Duplex mode
No network congestion discovered.
Good network cable(s) found
Normal duplex operation found.

Web100 reports the Round trip time = 440.81 msec; the Packet size = 1460 Bytes; and
No packet loss - but packets arrived out-of-order 1.52% of the time
This connection is receiver limited 76.22% of the time.
This connection is network limited 23.76% of the time.

Your ethernet link has been set to half-duplex. This means signals can only be sent in one direction through the cables at any one time. The normal setting should be full-duplex.

Latency (round trip time) is high at 441ms. Normal values should be 90-130ms. Data packets arriving out-of order.

What kind of devices and connections do you have from your comp to the splitter?

[Da3dlaus]

We have a D-Link DSS-5+ 10/100 Fst Ethernet Switch connected to the cable modem, but were not using a router. Ethernet cables are hooked up to the computers from there.

And how do i change those settings?

[trogers]

Check the duplex setting of your NIC.

Start=>Control Panel=>System=>Hardware=>Device Mgr=>Network Adapters

Click right on NIC and select 'Properties' and Advanced

Go to Speed and Duplex and select 'Auto-sense'

Click ok and turn off power to computer and unplug the cable from the switch to your comp for a few mins.

Restart comp and reconnect cable and do the Network Diagnostic test again to see if your link is set to full-duplex.

[Da3dlaus]

Fixed that, and now im seeing a higher latency number thatn i was before...heres teh results:

WEB100 Enabled Statistics:
Checking for Middleboxes . . . . . . . . . . . . . . . . . . Done
running 10s outbound test (client to server) . . . . . 372.54Kb/s
running 10s inbound test (server to client) . . . . . . 3.71Mb/s

------ Client System Details ------
OS data: Name = Windows XP, Architecture = x86, Version = 5.1
Java data: Vendor = Sun Microsystems Inc., Version = 1.5.0_05

------ Web100 Detailed Analysis ------
Cable modem/DSL/T1 link found.
Link set to Full Duplex mode
No network congestion discovered.
Good network cable(s) found
Normal duplex operation found.

Web100 reports the Round trip time = 429.88 msec; the Packet size = 1460 Bytes; and
No packet loss - but packets arrived out-of-order 3.2% of the time
This connection is receiver limited 56.87% of the time.
This connection is network limited 43.1% of the time.

Web100 reports TCP negotiated the optional Performance Settings to:
RFC 2018 Selective Acknowledgment: ON
RFC 896 Nagle Algorithm: ON
RFC 3168 Explicit Congestion Notification: OFF
RFC 1323 Time Stamping: OFF
RFC 1323 Window Scaling: ON
Packet size is preserved End-to-End
Server IP addresses are preserved End-to-End
Client IP addresses are preserved End-to-End

[trogers]

Web100 reports the Round trip time = 429.88 msec; the Packet size = 1460 Bytes; and
No packet loss - but packets arrived out-of-order 3.2% of the time
This connection is receiver limited 56.87% of the time.
This connection is network limited 43.1% of the time.

Latency (round trip time) is high at 430ms. Normal range should be 90-130ms.

Check and ensure there is no other electrical device placed near to the cable modem or the switch. Also, the modem and the switch should be kept a few feet apart.

Do the network test again when the other comp(s) are not in use.

[Da3dlaus]

hmm...the modem is right on top of one computer, and the switch is nestled in with the other...xD

Ok, iv repositioned both the modem and the switch. The second computer is turned off, and ive put in a brand new coaxial cable (other one was old)

And thanks for being so patient with me ^^

[trogers]

Do the network diagnostic test again to see if your latency has fallen.

[Da3dlaus]

Still getting high lantency times, and at this point, im clueless.

WEB100 Enabled Statistics:
Checking for Middleboxes . . . . . . . . . . . . . . . . . . Done
running 10s outbound test (client to server) . . . . . 378.33Kb/s
running 10s inbound test (server to client) . . . . . . 3.94Mb/s

------ Client System Details ------
OS data: Name = Windows XP, Architecture = x86, Version = 5.1
Java data: Vendor = Sun Microsystems Inc., Version = 1.5.0_05

------ Web100 Detailed Analysis ------
Cable modem/DSL/T1 link found.
Link set to Half Duplex mode
No network congestion discovered.
Good network cable(s) found
Normal duplex operation found.

Web100 reports the Round trip time = 437.57 msec; the Packet size = 1460 Bytes; and
No packet loss - but packets arrived out-of-order 3.03% of the time
This connection is receiver limited 66.64% of the time.
This connection is network limited 33.33% of the time.

Web100 reports TCP negotiated the optional Performance Settings to:
RFC 2018 Selective Acknowledgment: ON
RFC 896 Nagle Algorithm: ON
RFC 3168 Explicit Congestion Notification: OFF
RFC 1323 Time Stamping: OFF
RFC 1323 Window Scaling: ON
Packet size is preserved End-to-End
Server IP addresses are preserved End-to-End
Client IP addresses are preserved End-to-End

this is with a fresh restart, too oO

[Da3dlaus]

wtf? duplex reset itself...and i have the same settings for the NIC, totally unchanged

[Da3dlaus]

another test, and now its saying i have 2 different duplex settings. However, my latency is now down to 335 seconds.
WEB100 Enabled Statistics:
Checking for Middleboxes . . . . . . . . . . . . . . . . . . Done
running 10s outbound test (client to server) . . . . . 340.78Kb/s
running 10s inbound test (server to client) . . . . . . 1.74Mb/s

------ Client System Details ------
OS data: Name = Windows XP, Architecture = x86, Version = 5.1
Java data: Vendor = Sun Microsystems Inc., Version = 1.5.0_05

------ Web100 Detailed Analysis ------
Cable modem/DSL/T1 link found.
Link set to Full Duplex mode
No network congestion discovered.
Good network cable(s) found
Alarm: Duplex mismatch condition found: Host set to Full and Switch set to Half duplexD

Web100 reports the Round trip time = 335.0 msec; the Packet size = 1460 Bytes; and
No packet loss - but packets arrived out-of-order 14.79% of the time
This connection is receiver limited 3.74% of the time.
This connection is network limited 96.23% of the time.

Web100 reports TCP negotiated the optional Performance Settings to:
RFC 2018 Selective Acknowledgment: ON
RFC 896 Nagle Algorithm: ON
RFC 3168 Explicit Congestion Notification: OFF
RFC 1323 Time Stamping: OFF
RFC 1323 Window Scaling: ON
Packet size is preserved End-to-End
Server IP addresses are preserved End-to-End
Client IP addresses are preserved End-to-End

[trogers]

You will need to do some investigative work to isolate the cause of the changing duplex mode, high latency and data packets arriving out of order.

To do this, you can use the elimination method.

First connect your comp directly to the cable modem and do the network diagnostic test and print out your results. Next do the same direct connection with the other comp and print out the results.

Then connect each of the 2 comps to the switch, one at a time, and test again. Compare all the test results.

If using the switch gives both comp the high latency, you know that is the cause. If only one comp gets the high latency, the problem may be in the cable or the NIC of that comp.

[Da3dlaus]

I got similar results on the first computer, bypassing and using the switch. Both were above 400ms latency. The other computer gave teh same results, both with and without the switch.

At this point im guessing it could be a weak signal in the cabel line, as that was the reason we could not even connect back in July. Someone from the cable company had to come out and fix it. He said something about running the cable through an amplifier degrades or wekaens the signal...?

[trogers]

Your conclusion is probably correct. If both comps show high latency when connected directly to the cable modem, the cause of the problem is from the modem to the ISP box outside the house.