dslreports logo
site
 
    All Forums Hot Topics Gallery
spc

spacer




how-to block ads


Search Topic:
uniqs
1616
share rss forum feed

AlmightyOne

join:2003-03-02
Riverview, FL

2 edits

[Southeast] 150/35 Questions

So I just got this service installed and have a few questions. I will start off with some of the results I have got so far:

Verizon Speed Test

Checking for Middleboxes . . . . . . . . . . . . . . . . . . Done
SendBufferSize set to [261360]
running 10s outbound test (client to server) . . . . . 75.21Mb/s
running 10s inbound test (server to client) . . . . . . 153.31Mb/s

------ Client System Details ------
OS data: Name = Windows 7, Architecture = x86, Version = 6.1
Java data: Vendor = Sun Microsystems Inc., Version = 1.6.0_31

------ Web100 Detailed Analysis ------
Client Receive Window detected at 261340 bytes.
10 Gbps 10 GigEthernet/OC-192 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 = 12.36 msec; the Packet size = 1460 Bytes; and
No packet loss - but packets arrived out-of-order 0.04% of the time
This connection is receiver limited 3.47% of the time.
Increasing the the client's receive buffer (255.0 KB) will improve performance
This connection is sender limited 96.23% of the time.




So based off these speedtests, I see that I am provisioned for 160/75 based on the Verizon Speedtest. Now Tampa has always been crappy when it comes to exterior speed tests and the peering agreements the ISP's have forced routing to start outside of Tampa (ie. Atlanta or Miami) is usually our first hop out. My question is do you think that the speedtest results are severely limiting the bandwidth on those results? Anyone else that is in the Southeast Region have 150/35 that could test or share results.

The second interesting thing I found that was after running the Verizon TCP Optimizer, that my MTU was set to 1300. I have to ask why? Here is what I found and what I did to change it:

C:\>netsh interface ipv4 show subinterfaces

MTU MediaSenseState Bytes In Bytes Out Interface
------ --------------- --------- --------- -------------
1492 1 0 19330 Loopback Pseudo-Interface 1
1300 5 0 0 Wireless Network Connection
1300 1 221154642 94069205 Local Area Connection

Changed it by entering the following:

C:\>netsh interface ipv4 set subinterface "Local Area Connection" mtu=1500 store=persistent
C:\>netsh interface ipv4 set subinterface "Wireless Network Connection" mtu=1500 store=persistent
C:\>netsh interface ipv4 show subinterfaces

MTU MediaSenseState Bytes In Bytes Out Interface
------ --------------- --------- --------- -------------
1492 1 0 19330 Loopback Pseudo-Interface 1
1500 5 0 0 Wireless Network Connection
1500 1 221154642 94069205 Local Area Connection

Can others who have the 150/35 verify what their MTU settings are? Also did you run the Verizon Optimizer or other methods for adjusting your MTU and Rx Window Sizes? I am basing this off of Windows 7.

Also, thought I would share a quick tracert to dslreports.com and google.com although I am quite happy with the latency.

Tracing route to www.dslreports.com [209.123.109.175]
over a maximum of 30 hops:

1 1 ms 1 ms 1 ms 192.168.1.1
2 1 ms 3 ms 1 ms L100.TAMPFL-VFTTP-71.verizon-gni.net [173.65.173.1]
3 4 ms 5 ms 5 ms G0-5-4-4.TAMPFL-LCR-21.verizon-gni.net [130.81.109.244]
4 5 ms 5 ms 5 ms 130.81.199.26
5 18 ms 18 ms 18 ms ge-5-2-0-0.ATL01-BB-RTR1.verizon-gni.net [130.81.17.115]
6 17 ms 18 ms 18 ms 0.ae3.BR3.ATL4.ALTER.NET [152.63.80.73]
7 20 ms 20 ms 20 ms xe-9-0-0.edge4.Atlanta2.Level3.net [4.68.62.121]
8 32 ms 23 ms 24 ms vlan52.ebr2.Atlanta2.Level3.net [4.69.150.126]
9 20 ms 22 ms 20 ms ae-73-73.ebr3.Atlanta2.Level3.net [4.69.148.253]
10 36 ms 33 ms 33 ms ae-2-2.ebr1.Washington1.Level3.net [4.69.132.86]
11 46 ms 35 ms 39 ms ae-71-71.csw2.Washington1.Level3.net [4.69.134.134]
12 34 ms 33 ms 33 ms ae-72-72.ebr2.Washington1.Level3.net [4.69.134.149]
13 38 ms 45 ms 49 ms ae-4-4.ebr2.Newark1.Level3.net [4.69.132.102]
14 39 ms 39 ms 37 ms ae-21-52.car1.Newark1.Level3.net [4.69.156.37]
15 40 ms 40 ms 41 ms NETCCESS.car1.Newark1.Level3.net [4.26.16.190]
16 49 ms 49 ms 49 ms 0.e3-3.tbr2.mmu.nac.net [209.123.11.77]
17 44 ms 44 ms 44 ms 0.e1-1.tbr2.oct.nac.net [209.123.10.21]
18 45 ms 44 ms 44 ms vlan808.esd1.oct.nac.net [209.123.10.42]
19 41 ms 40 ms 41 ms www.dslreports.com [209.123.109.175]

Tracing route to www.l.google.com [173.194.37.82]
over a maximum of 30 hops:

1 1 ms 1 ms 1 ms 192.168.1.1
2 1 ms 1 ms 1 ms L100.TAMPFL-VFTTP-71.verizon-gni.net [173.65.173.1]
3 3 ms 3 ms 3 ms G0-5-4-4.TAMPFL-LCR-22.verizon-gni.net [130.81.110.218]
4 5 ms 5 ms 5 ms so-2-0-0-0.TPA01-BB-RTR2.verizon-gni.net [130.81.28.214]
5 12 ms 12 ms 14 ms 0.xe-7-1-0.XL2.MIA19.ALTER.NET [152.63.5.185]
6 12 ms 10 ms 10 ms 0.xe-11-2-0.GW1.MIA19.ALTER.NET [152.63.80.98]
7 37 ms 10 ms 11 ms google-gw.customer.alter.net [63.65.188.106]
8 10 ms 10 ms 11 ms 209.85.253.120
9 24 ms 24 ms 22 ms 216.239.48.192
10 24 ms 24 ms 24 ms 64.233.175.94
11 23 ms 24 ms 24 ms atl14s08-in-f18.1e100.net [173.194.37.82]

My last comment is in regards to Wireless. I received an Actiontec Rev. I router with my installation. My results are extremely sporadic when it came to Wireless but I also don't have any devices that have the newest "N" but rather are draft N. Could others please post their results if you have the 150 package and what the best Wireless speedtests you could get?


houkouonchi

join:2002-07-22
Ontario, CA
Reviews:
·Verizon FiOS
said by AlmightyOne:

------ Web100 Detailed Analysis ------
Client Receive Window detected at 261340 bytes.
10 Gbps 10 GigEthernet/OC-192 link found.
Link set to Full Duplex mode
No network congestion discovered.
Good network cable(s) found
Normal duplex operation found.

You are lucky you are getting 150 megabits even on the vz speed test with such a low receive window buffer (250 kilobytes). It should be atleast 1 megabyte for the 150 meg service and even higher. It could also be that speedtest.net is not on a 1gbit ocnnection. I know most of the major areas are but not all of the servers are.

As for wireless especially draft N I was lucky to get 120-130 meg real-world results when I was sitting right on top of the darn thing. At normal distances i would be in the 50-80 megabit range compared to G's 10-20 megabit range.
--
300/150 mbit Bonded Verizon FiOS connection FTW!

AlmightyOne

join:2003-03-02
Riverview, FL

4 edits

You are lucky you are getting 150 megabits even on the vz speed test with such a low receive window buffer (250 kilobytes). It should be atleast 1 megabyte for the 150 meg service and even higher. It could also be that speedtest.net is not on a 1gbit ocnnection. I know most of the major areas are but not all of the servers are.

Well I just did some research, and I should have been more specific with some of the items I stated. First off I am running Windows 7 and based on my reading it uses autotuning values for RWIN. I assume that you are statically entering your RWIN value in. What OS are you using? For Windows 7 users, how would you disable autotuning RWIN and statically enter it (Registry?, Programs?, etc). Would the benefits outweigh the downfalls is you use a static method vs autotuning due to delay always being different depending on where you are downloading/uploading from?

I just tried SG TCPOptimizer but I see no place that you can manually set the RWIN and where you disable autotuning.

Here is an example of running the Verizon Speed Test to California where you can see the RWIN is significantly higher:

Analysis information:

Checking for Middleboxes . . . . . . . . . . . . . . . . . . Done
SendBufferSize set to [261360]
running 10s outbound test (client to server) . . . . . 27.29Mb/s
running 10s inbound test (server to client) . . . . . . 140.44Mb/s

------ Client System Details ------
OS data: Name = Windows 7, Architecture = x86, Version = 6.1
Java data: Vendor = Sun Microsystems Inc., Version = 1.6.0_31

------ Web100 Detailed Analysis ------
Client Receive Window detected at 6334720 bytes.
622 Mbps OC-12 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 = 309.78 msec; the Packet size = 1460 Bytes; and
No packet loss - but packets arrived out-of-order 0.01% of the time
This connection is receiver limited 2.21% of the time.
Increasing the the client's receive buffer (6186.0 KB) will improve performance
This connection is sender limited 93.84% of the time.
This connection is network limited 3.95% 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
Information: Network Address Translation (NAT) box is modifying the Client's IP address

Another interesting tidbit I found was that after rebooting my computer to adjust the settings from SG TCP Optimizer, that my MTU values defaulted back to 1300. Does anyone know how to keep the MTU adjusted at 1500 after a reboot with Windows 7. I remember having this problem before with my old internet package but the MTU value was 1492 and I would have to manually enter in 1500 through DOS prompt to fix it everytime. After awhile I gave up and forgot to ever fix it.

I edited my first post to reflect that information.


Smith6612
Premium,MVM
join:2008-02-01
North Tonawanda, NY
kudos:24
Reviews:
·Verizon Online DSL
·Frontier Communi..
Do you have auto-tuning Heuristics enabled by any chance on your system? That was introduced starting with Windows Vista Service Pack 1 I believe and it may be adjusting the settings like that. Try disabling it if it's on.

netsh int tcp set heuristics disabled

If it's on can be shown through

netsh int tcp show heuristics

AlmightyOne

join:2003-03-02
Riverview, FL
said by Smith6612:

Do you have auto-tuning Heuristics enabled by any chance on your system? That was introduced starting with Windows Vista Service Pack 1 I believe and it may be adjusting the settings like that. Try disabling it if it's on.

netsh int tcp set heuristics disabled

If it's on can be shown through

netsh int tcp show heuristics

Yes with SG TCP Optimizer it gave me the option to disable it. I verified with your command in DOS prompt and it shows it is disabled.


jarablue
Always be true to yourself

join:2001-06-11
Boxborough, MA
reply to AlmightyOne
What are the tweaks needed for Windows 7 64 bit on 150meg?

I am thinking about subscribing when the 17th hits and would like my OS ready for the speed.

serge87

join:2009-11-29
Reviews:
·Verizon FiOS
said by jarablue:

What are the tweaks needed for Windows 7 64 bit on 150meg?

I am thinking about subscribing when the 17th hits and would like my OS ready for the speed.

I personally use these settings and get my full 150/65:

quote:
netsh interface tcp show global
Querying active state...
TCP Global Parameters
----------------------------------------------
Receive-Side Scaling State : enabled
Chimney Offload State : disabled
NetDMA State : enabled
Direct Cache Acess (DCA) : enabled
Receive Window Auto-Tuning Level : experimental
Add-On Congestion Control Provider : ctcp
ECN Capability : disabled
RFC 1323 Timestamps : disabled
In my opinion, I wouldn't even bother using the Verizon "speed optimizer", it's a pathetic joke. Use at your own risk.

When the tech came to upgrade me to 150/35 from 50/20, he spent an hour trying to get my upload speed from 10mbps to the 65mbps. I had the Rev. G Actiontec in place coupled with a new server-variant Intel gigabit NIC. After running the VZ optimizer for the fourth or fifth time he gave up and was about to leave when I asked if I could revert to my saved TCP/IP settings(always keep a backup!) to pre-"VZ optimization" state and that did the trick.

In addition to the settings I posted above, I also had to tune my 'interrupt moderation rate' to maximum on my NIC to receive highest throughput. Your NIC may or may not have this option.

A short explanation:

quote:
Interrupt Moderation Rate

This sets the rate at which the controller moderates or delays the generation of interrupts making it possible to optimize network throughput and CPU utilization. The default setting (Adaptive) adjusts the interrupt rates dynamically depending on traffic type and network usage. Choosing a different setting may improve network and system performance in certain configurations.

Without interrupt moderation, CPU utilization increases at higher data rates because the system must handle a larger number of interrupts. Interrupt moderation causes the network driver to accumulate interrupts and send a single interrupt rather than a series of interrupts. At higher data rates, a high interrupt moderation setting may improve system performance. At low data rates, a lower interrupt moderation setting is preferred as delayed interrupts causes latency.


jarablue
Always be true to yourself

join:2001-06-11
Boxborough, MA

2 edits
reply to AlmightyOne
Thanks for the help.


jarablue
Always be true to yourself

join:2001-06-11
Boxborough, MA

1 edit
reply to AlmightyOne
I've run the tcp optimizer and wanted to know if these settings are ok. Running 150/65 with W7 64bit 12gigs ram i7 920 cpu.

« SpeedGuide.net TCP Analyzer Results »
Tested on: 2012.06.27 02:02
IP address: 108.49.xx.xxx
Client OS/browser: Windows 7 (Internet Explorer 9.0)

TCP options string: 020405b40103030201010402
MSS: 1460
MTU: 1500
TCP Window: 65700 (multiple of MSS)
RWIN Scaling: 2 bits (2^2=4)
Unscaled RWIN : 16425
Recommended RWINs: 64240, 128480, 256960, 513920, 1027840
BDP limit (200ms): 2628kbps (329KBytes/s)
BDP limit (500ms): 1051kbps (131KBytes/s)
MTU Discovery: ON
TTL: 49
Timestamps: OFF
SACKs: ON
IP ToS: 00000000 (0)