All of lore.kernel.org
 help / color / mirror / Atom feed
From: Rani Ahmed <rani79@idm.net.lb>
To: lartc@vger.kernel.org
Subject: [LARTC] problem in speeds [Message from superlinux]
Date: Mon, 23 May 2005 07:06:03 +0000	[thread overview]
Message-ID: <429180DB.9060000@idm.net.lb> (raw)

i am assigned a network to  replace its "Windows server with ISA caching 
proxy" with another "debian linux with squid proxy" with both  "linux" 
and "ISA" are completely differnet boxes. i am using  linux 2.6 kernel 
since the linux server has SATA hard disks .

the network has downlink with a penta@net DVB card for down-link ; then 
it's connected to the ISA server which   distributes the connections to 
3 other networks. the UP-Link is routed thru a cisco router which is 
connected with an ISP selling the network owners  the bandwith of 
512Kbit/sec DownLink and 128kbps UPLink. this is a sketch  which might  
can help:


                                                                         
                                                downlink          LAN1 
                                                                         
                                                 ----->                 
/\   
Internet  >>> PENTA@NET on Windows server ] >>> [ISA proxy server on 
Windows] ===>>LAN2
                                                                        
                        ||                                               
   \/  
                                                                         
                       ||                                              
LAN3        
               Internet <<<<ISP<<<UPlink <<<<Cisco router||
                                                                         
                          
what i have done on linux is just a simple NAT and a redirect to Squid 
to force transparent cache so the user does not have to configure his 
browser.

The Problem:
=====

I did not do any shaping, even on squid no shaping, and all i get using 
the NETSPEED GNOME plugin on the Internet side is 16 Kbytes per sec up 
and 32Kbytes/sec on the Local side as total speed of the whole 3 
networks.  and as you should have noticed, my Linux box should work as a 
water hause or pipe and has no real effect on the bandwidth.

if i returned back the windows ISA box the speed goes normally at its 
52kbyte/sec.

so what is happening FOR GOD'S SAKE!!???

in addition , after all that,  i thought it was slow because of the RCV 
and SEND buffers settings in the kernel an the TTLd. so set RCV and 
SEND as twice the kernel's default on /etc/sysctl.conf as:

net.core.rmem_default = 262140
net.core.wmem_default = 262140
net.core.rmem_max = 262140
net.core.wmem_max = 262140

the RESULT IS THE SAME!!!

so what is happening FOR GOD'S SAKE!!???
 
_______________________________________________
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc

                 reply	other threads:[~2005-05-23  7:06 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=429180DB.9060000@idm.net.lb \
    --to=rani79@idm.net.lb \
    --cc=lartc@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body. 
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.