Re: scp and sftp: large files not being transfered

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Saw your earlier question but didn't notice that you were on a wireless
device.  Don't know if it's related, but in my experience large packets +
wireless = bad.  I think (not to get off topic here) that this is because
TCP makes assumptions about the meaning of lost packets (namely that
they're CONGESTION related).  I've seen radio links where TCP will back
off to a complete standstill, but only in the upload direction, downloads
can work at 4Mbps.  What matters here is the BER, and since larger packets
have more bits in them, the PER goes up for larger packets, which are the
usual case in uploads, hence a larger fraction of packets is lost when
uploading than when downloading, and TCP goes crazy.  Can you isolate the
wireless device here, and prove that it A)IS or B)IS NOT a problem
involving wireless?

Are there other values of MTU that you've tried?  What happens if you lock
MTU at 1500, or it's close relatives, like 1450, and so on?

On Wed, 22 Sep 2004, nikhil hegde wrote:

> Date: Wed, 22 Sep 2004 11:04:49 -0700 (PDT)
> From: nikhil hegde <ndhegde@xxxxxxxxx>
> To: netfilter@xxxxxxxxxxxxxxxxxxx
> Subject: scp and sftp: large files not being transfered
>
> Hi All,
> I've mentioned this prblem earlier. It is about large
> file transfer not taking place in fedora core 2 kernel
> 2.6.5-1.358. When I am getting packets into userspace
> and doing some processing.
>
> http://lists.netfilter.org/pipermail/netfilter/2004-September/056036.html
>
> I was doing "ip route change 10.6/16 mtu 1000 dev
> wlan0"
>
> If I lock the mtu value then it just works fine. Does
> it mean that this kernel is not considering my request
> to set the mtu to 1000?because..I read that..
>
> "If the modifier lock is not used, the MTU may be
> updated by the kernel due to Path MTU Discovery. If
> the modifier lock is used, no path MTU discovery will
> be tried, all packets will be sent without the DF bit
> in IPv4 case or fragmented to MTU for IPv6"
>
> If this is so what is the role of
> "/proc/sys/net/ipv4/ip_no_pmtu_disc " , which still
> shows the value 0. Also given the fact that transfer
> happens on 2.4.18-3 kernel with same pmtu setting.
>
> Could anyone pleasse explain me the fundaes behind
> this?
>
> thanks
> Nikhil
>
>
>
>
>
>
>
> __________________________________
> Do you Yahoo!?
> New and Improved Yahoo! Mail - 100MB free storage!
> http://promotions.yahoo.com/new_mail
>


[Index of Archives]     [Linux Netfilter Development]     [Linux Kernel Networking Development]     [Netem]     [Berkeley Packet Filter]     [Linux Kernel Development]     [Advanced Routing & Traffice Control]     [Bugtraq]

  Powered by Linux