The receive buffer ?

stumpster · Post by **stumpster** » Fri Oct 06, 2006 3:42 am

Hi, I ran the test that u recommended others and got this, but I don't understand what its telling me to do?(the highlighted parts)
I would greatly appreciate your help!
Information: The receive buffer should be 71.50 Kbytes to maximize throughput

WEB100 Enabled Statistics:
Checking for Middleboxes . . . . . . . . . . . . . . . . . . Done
running 10s outbound test (client to server) . . . . . 219.39Kb/s
running 10s inbound test (server to client) . . . . . . 1.29Mb/s

------ Client System Details ------
OS data: Name = Windows XP, Architecture = x86, Version = 5.1
Java data: Vendor = Sun Microsystems Inc., Version = 1.5.0_06

------ 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 = 195.26 msec; the Packet size = 1440 Bytes; and
No packet loss was observed.
This connection is receiver limited 93.85% of the time.
Increasing the the client's receive buffer (30.0 KB) will improve performance
Increasing the the client's receive buffer (30.0 KB) will improve performance
This connection is network limited 6.13% 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: OFF
Information: Network Middlebox is modifying MSS variable
Server IP addresses are preserved End-to-End
Client IP addresses are preserved End-to-End

stumpster · Post by **stumpster** » Fri Oct 06, 2006 3:44 am

Oh, and here is the other!
TCP options string = 020405a001010402

MTU = 1480
MTU is optimized for Windows XP PPoE DSL broadband. If not, consider raising MTU to 1500 for optimal throughput.

MSS = 1440
MSS is optimized for Windows XP PPPoE DSL broadband. If not, consider raising your MTU value.

Default TCP Receive Window (RWIN) = 31680 This is the only thing I get when I search receive buffer. Is that what its referring to?
RWIN Scaling (RFC1323) = 0 bits
Unscaled TCP Receive Window = 31680

RWIN is a multiple of MSS
Other RWIN values that might work well with your current MTU/MSS:
506880 (MSS x 44 * scale factor of 8)
253440 (MSS x 44 * scale factor of 4)
126720 (MSS x 44 * scale factor of 2)
63360 (MSS x 44)

bandwidth * delay product (Note this is not a speed test):

Your TCP Window limits you to: 1267.2 kbps (158.4 KBytes/s) @ 200ms
Your TCP Window limits you to: 506.88 kbps (63.36 KBytes/s) @ 500ms

MTU Discovery (RFC1191) = ON

Time to live left = 54 hops
TTL value is ok.

Timestamps (RFC1323) = OFF
Selective Acknowledgements (RFC2018) = ON

IP type of service field (RFC1349) = 00000000 (0)

Thank you !

trogers · Post by **trogers** » Fri Oct 06, 2006 3:53 am

The TCP receive buffer in the Network Diagnostic test is the same as the TCP Receive Window (RWIN) in the TCP Analyzer. It is measured in KiloBytes (KB).

Your present buffer size is about 30 KB (31680) and the test recommends to make it 71.50 KB.

We use the TCP Optimizer to fine tune the correct buffer size or RWIN.

To do this, we need to know your subscribed speed, your location (country) and your operating system. We can only tweak Windows OS here.

stumpster · Post by **stumpster** » Fri Oct 06, 2006 3:58 am

Ok, I have 1.5 Mbps I'm located in the US Mid west and have windows XP!
Thanks a lot I didn't expect a reply so soon!

trogers · Post by **trogers** » Fri Oct 06, 2006 4:06 am

If you are in the US, your RWIN is already in the normal range for cable and DSL broadband.

What is not normal is your latency (Round trip time) - 195ms. The usual value ranges between 90-130ms. It is due to this high latency the test recommends buffer size should be increased.

Your signal quality seems to be good. A tracert may help show the cause for the high latency.

To do a tracert, click 'Start' and then 'Run'. Type "cmd" into the box and click 'ok'.

A DOS black screen will appear. Type after the C:\

tracert http://www.speedguide.net

and press enter key

When test is completed, move mouse cusor to the DOS black screen, click right and choose 'select all'.

Move mouse cursor to the Quick Reply box of this forum, click right and 'paste'.

stumpster · Post by **stumpster** » Fri Oct 06, 2006 4:11 am

Microsoft Windows XP [Version 5.1.2600]
(C) Copyright 1985-2001 Microsoft Corp.

C:\Documents and Settings\mark>tracert http://www.speedguide.net

Tracing route to speedguide.net [63.217.30.70]
over a maximum of 30 hops:

1 19 ms 19 ms 18 ms eaglcocobr1.lajncocoro2.centurytel.net [69.xx.xxxxxxxx
2 19 ms 21 ms 21 ms GE-0-0-0-j1-rt.eag.centurytel.net [207.xxx.xxxx
]
3 27 ms 26 ms 26 ms so-10-3.hsa1.Denver1.Level3.net [209.245.23.121]

4 27 ms 27 ms 24 ms ae-1-53.bbr1.Denver1.Level3.net [4.68.107.65]
5 38 ms 40 ms 39 ms ae-0-0.bbr2.Dallas1.Level3.net [64.159.1.110]
6 41 ms 75 ms 41 ms ae-24-56.car4.Dallas1.Level3.net [4.68.122.176]

7 41 ms 40 ms 41 ms savvis-level3-ge.Dallas1.Level3.net [4.68.111.86
]
8 39 ms 41 ms 75 ms dpr1-ge-0-3-1.dallasequinix.savvis.net [204.70.1
94.21]
9 40 ms 41 ms 40 ms 208.173.178.130
10 73 ms 71 ms 70 ms pos5-1.br01.ash01.pccwbtn.net [63.218.44.54]
11 73 ms 72 ms 72 ms 63-217-30-70.pccwbtn.net [63.217.30.70]

Trace complete.

C:\Documents and Settings\mark>

trogers · Post by **trogers** » Fri Oct 06, 2006 4:17 am

stumpster wrote: 1 19 ms 19 ms 18 ms eaglcocobr1.lajncocoro2.centurytel.net [69.29.xx
x.xxx]

We usually x out the last 6 digits of the first and second hops to prevent hackers knowing our IP address.

The time value 19ms is rather high. Usual value is 9-11ms. Check and ensure your modem is not placed near to other electrical device such as power box, speakers, cordless phone, etc. If yes, relocate the modem a few feet away, power off the modem for a few mins and then reconnect.

Do the Network Diagnostic test again to see if your latency falls.

stumpster · Post by **stumpster** » Fri Oct 06, 2006 4:22 am

We usually x out the last 6 digits of the first and second hops to prevent hackers knowing our IP address.
It wont let me edit it now?
I'm moving my modem now!

trogers · Post by **trogers** » Fri Oct 06, 2006 4:25 am

That is ok. Your IP address will usually change after power off and reconnecting your modem.

stumpster · Post by **stumpster** » Fri Oct 06, 2006 4:48 am

Ok sorry it took so long. This is what I got!
WEB100 Enabled Statistics:
Checking for Middleboxes . . . . . . . . . . . . . . . . . . Done
running 10s outbound test (client to server) . . . . . 196.70Kb/s
running 10s inbound test (server to client) . . . . . . 1.28Mb/s

------ Client System Details ------
OS data: Name = Windows XP, Architecture = x86, Version = 5.1
Java data: Vendor = Sun Microsystems Inc., Version = 1.5.0_06

------ 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 = 194.88 msec; the Packet size = 1440 Bytes; and
No packet loss was observed.
This connection is receiver limited 93.66% of the time.
Increasing the the client's receive buffer (30.0 KB) will improve performance
This connection is network limited 6.31% 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: OFF
Information: Network Middlebox is modifying MSS variable
Server IP addresses are preserved End-to-End
Client IP addresses are preserved End-to-End

No change! Tracert seems to be worse?

1 18 ms 31 ms 31 ms eaglcocobr1.lajncocoro2.centurytel.net [69.29.xxxxxx]
2 31 ms 31 ms 31 ms GE-0-0-0-j1-rt.eag.centurytel.net [207.1xxxxxxxxxx
]
3 31 ms 31 ms 31 ms so-10-3.hsa1.Denver1.Level3.net [209.245.23.121]

4 31 ms 31 ms 31 ms ae-1-55.bbr1.Denver1.Level3.net [4.68.107.129]
5 46 ms 46 ms 46 ms ae-0-0.bbr2.Dallas1.Level3.net [64.159.1.110]
6 46 ms 46 ms 46 ms ae-24-56.car4.Dallas1.Level3.net [4.68.122.176]

7 46 ms 46 ms 46 ms savvis-level3-ge.Dallas1.Level3.net [4.68.111.90
]
8 46 ms 46 ms 46 ms dpr1-ge-1-3-0.dallasequinix.savvis.net [204.70.1
94.13]
9 46 ms 46 ms 46 ms 208.173.178.130
10 78 ms 78 ms 78 ms pos5-1.br01.ash01.pccwbtn.net [63.218.44.54]
11 78 ms 78 ms 78 ms 63-217-30-70.pccwbtn.net [63.217.30.70]

Trace complete.

trogers · Post by **trogers** » Fri Oct 06, 2006 4:57 am

Seems there is no change. If you can, borrow a modem from a friend to test your connection latency. If values falls to normal range, then you may need to replace your modem. For now we increase your buffer (RWIN) to compensate for the high latency.

Use the TCP Optimizer with these custom settings for optimizing:

General Settings tab:
Custom settings - check
Modify All Network Adapters - check
network adapter selection - your NIC
MTU - 1480
TTL - 64
TCP Receive Window - 63360
MTU Discovery - Yes
Black Hole Detect - No
Selective Acks - Yes
Max Duplicate ACKs - 2
TCP 1323 Options:
Windows Scaling - uncheck
Timestamps - uncheck
Advanced Settings tab:
Max Connections per Server - 10
Max Connections per 1.0 Server - 20
LocalPriority - 5
Host Priority - 6
DNSPriority - 7
NetbtPriority - 8
Lan Browsing speedup - optimized
QoS: NonBestEffortLimit - 0
ToS: DisableUserTOSSetting - 0
ToS: DefaultTOSValue - 80
MaxNegativeCacheTtl - 0
NetFailureCacheTime - 0
NegativeSOACache Time - 0
LAN Request Buffer Size - 32768
Then select "Apply Changes" and reboot to take effect

stumpster · Post by **stumpster** » Fri Oct 06, 2006 5:11 am

Ok this is with changes applied!
WEB100 Enabled Statistics:
Checking for Middleboxes . . . . . . . . . . . . . . . . . . Done
running 10s outbound test (client to server) . . . . . 187.13Kb/s
running 10s inbound test (server to client) . . . . . . 1.27Mb/s

------ Client System Details ------
OS data: Name = Windows XP, Architecture = x86, Version = 5.1
Java data: Vendor = Sun Microsystems Inc., Version = 1.5.0_06

------ 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 = 362.78 msec; the Packet size = 1440 Bytes; and
There were 5 packets retransmitted, 40 duplicate acks received, and 42 SACK blocks received
The connection was idle 0 seconds (0%) of the time
This connection is receiver limited 82.27% of the time.
Increasing the the client's receive buffer (63.0 KB) will improve performance
This connection is network limited 17.7% 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: OFF
Information: Network Middlebox is modifying MSS variable
Server IP addresses are preserved End-to-End
Client IP addresses are preserved End-to-End

trogers · Post by **trogers** » Fri Oct 06, 2006 5:20 am

[quote="stumpster"]
Web100 reports the Round trip time = 362.78 msec]

Now your latency nearly doubled. This is a sign that some source of electromagnetic interference is affecting your modem.

One way to isolate this source is to turn off all electrical devices in the room except your computer and modem. Test your latency and then turn the other electrical devices back on one at a time and test each time. This should incude the heat pump if you have any.

stumpster · Post by **stumpster** » Fri Oct 06, 2006 5:25 am

could it be interference with the phone line also, or just modem?

trogers · Post by **trogers** » Fri Oct 06, 2006 5:31 am

It could be if your one splitter is not connected directly to a wall jack or your cables are too long. Standard lengths are abour 5 feet.

stumpster · Post by **stumpster** » Fri Oct 06, 2006 5:33 am

the reason I ask is because I use surge protector with a splitter! So my phone line is right by all my cords. I just did again with all off accept computer, monitor and modem and this is result!
Cable modem/DSL/T1 link found.
Link set to Full Duplex mode
Information: throughput is limited by other network traffic.
Good network cable(s) found
Normal duplex operation found.

Web100 reports the Round trip time = 268.89 msec; the Packet size = 1440 Bytes; and
There were 7 packets retransmitted, 32 duplicate acks received, and 35 SACK blocks received
The connection was idle 0 seconds (0%) of the time
This connection is network limited 99.97% 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: OFF
Information: Network Middlebox is modifying MSS variable
Server IP addresses are preserved End-to-End
Client IP addresses are preserved End-to-End

stumpster · Post by **stumpster** » Fri Oct 06, 2006 5:45 am

Later on today I will run a designated line from outside to my modem with cat5 and see if that helps!
Thanks for all your time and help!

trogers · Post by **trogers** » Fri Oct 06, 2006 5:48 am

Ensure your splitter is not placed near to the power extension sockets. The fall in latency seems to be related to the load on your power sockets?

Turn on your other electrical device one at a time and test to see how latency relates to the device or power load.

Are you using a cordless phone set connected to your splitter?

stumpster · Post by **stumpster** » Fri Oct 06, 2006 5:59 am

No, I don't have cordless! My house is old though so if I run a new line maybe other then that I don't have any other power source by my modem! When u said it the first time I had my modem under my desk below the monitor and by my power strip!

My modem is like 6 years old also?

trogers · Post by **trogers** » Fri Oct 06, 2006 6:05 am

When you can bring down your latency to 90-130ms range, you can run the TCP Optimizer again with TCP Receive Window at 31680. Other fields remain the same.