• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

will jacking up MTU size increase speed

Red Squirrel

No Lifer
I have mostly gigabit throughout my home network. I find I just don't get the type of speeds gigabit should give me. Either I need to optimize stuff, or I'm simply getting what I paid for (consumer equipment vs commercial equipment).

I'm guessing the MTU on my windows machines and linux machines is around 1500. What if I was to set that to like say, 8000? Also, all machines on the same network segment need to have the same MTU right?

For a SMB transfer, I get about 107mb/sec (that's megabit). So at least I'm going over 100mb but not by much.
 
You can change it to the maximum the card and switch will allow and it will not harm TCP applications. SMB as a protocol is also notoriously slow.
 
Check your Internet bandwitdh at 1500, then increase it to 9000 and see how it is effecting Internet ,and LAN traffic (it usually does very little to local peer-to-peer traffic).

If it is No good roll it back, and No harm done.

What is probably more beneficial is changing the RCwin to a much larger number.

------------------
Use TCP Optimizer to optimize the TCP/IP stack of the computers according to the type and speed of the Internet connection.

http://www.speedguide.net/downloads.php

For 100Mb/sec. system use the recommended RCWin

For Giga Network double the recommended size of the RCWin.
 
Originally posted by: RedSquirrel
Hmm what is RCWin? Also what is a better protocol to test speed? Perhaps SMB is not the best test to use.

SMB is important as a "real world" test -- you should test as closely as you can to your intended usage. But for network tweaking, you should also test performance independently of the drives and file system, etc., using something like iperf 1.7:

server: iperf -s
client: iperf -c server -l 64k -t 15 -i 3 -r

where server is the name or IP of the remote computer running iperf -s.

http://www.noc.ucf.edu/Tools/Iperf/
 
Back
Top