Packet collision on Samba traffic
sharpe at ns.aus.com
Fri Apr 6 13:46:48 GMT 2001
At 03:37 PM 4/5/01 -0700, Andrew Tridgell wrote:
>> How about
>that looks really useful! Can you give us some figures for the
>improvement you see and with what NICs and client version?
>If this really does make a big difference then I wonder if we could
>automate it? Perhaps a libsmb based client testing against a share on
>the windows box could give a graph of speed versus SNDBUF?
Hmmm, I saw the following in the document:
>Actually, even in case of the other, more effective NIC, it should be better
>to set SO_SNDBUF size to multiple number of mss. Hence so, compare to
>SO_SNDBUF to 8192 byte, it is better to set it to 8576 bytes in case mss is
>536bytes ( since 8192/536 = 15.2...., 16*536 = 8576), and 8760 bytes in case
>mss is 1460 bytes ( 8192/1460 = 5.61... so 6 * 1460= 8760).
What is so interesting about this is that Windows NT sets the window size
to 8760, at least when communicating via Telnet. Actually, it first sets
the window size to 8192, then increases it to 8760.
Hmmm 2, I would be interested in writing a small application that would
probe this behaviour in Windows ...
Richard Sharpe, sharpe at ns.aus.com
Samba (Team member, www.samba.org), Ethereal (Team member, www.ethereal.com)
Contributing author, SAMS Teach Yourself Samba in 24 Hours
Author, Special Edition, Using Samba
More information about the samba-technical