mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Srihari Vijayaraghavan <harisri@bigpond.com>
To: lkml <linux-kernel@vger.kernel.org>
Subject: Request for help - tcpdump on many ethernet cards simulateneously
Date: Fri, 14 Mar 2003 00:51:19 +1100	[thread overview]
Message-ID: <200303140051.19453.harisri@bigpond.com> (raw)

Hello,

I have a requirement to capture network traffic on 5 Fast Ethernet cards 
simultaneously and store it in the local file system using tcpdump utility 
(3.7.2 Latest).

I ran some initial tests on RH 2.4.9 based kernel on a test machine with:
2 Xeon 2.8 GHz/512 KB Cache
1 GB RAM
U160 10K SCSI drives on Hardware RAID 1 under Compaq SmartArray controller 
(cciss.o)
4 Intel Ether Expro 100 Tx cards (eepro100.o), 1 Broadcom Gigabit (tg3.o)
All connected to a Cisco Fast Ethernet Switch (100 Tx only)

I captured approx 3 million packets of 1500 bytes on each adapter 
simultaneously over a period few minutes (it takes about 10 secs to fill up 
approx 500 MB in the EXT2 file system). During this period CPU (nearly 100% 
utilised), Memory (only few megabytes remained as free, rest all occupied by 
cache/buffer) and IO were really busy. The tcpdump utility reported that 
kernel hasn't dropped a single packet in that duration, which is a good news.

Is there anyone out there who has done similar work and would like to share 
the knowledge about:
1. Kernel version
2. File system used (parameters if any)
3. Network card and driver
4. SCSI/HW RAID controller card and driver
5. Tunning parameters for any sub-system if any
6. Any advise in general (don't use more than 1 GB RAM, use XFS, use aa/rmap, 
use 2.5 :-) etc..)

What I am really worried about is kernel may start dropping the packets after 
few hours/days and/or tcpdump/kernel may not be able to keep up with the 
network load due to IO load on the hard drives, memory pressure etc.. 

Are there any known bad effects on a 4 GB RAM configuration? (the production 
system will have 4 GB RAM)

By tomorrow I will have the opportunity to run it for few hours and see if it 
misbehaves (on 2.4.18-RH-latest and 2.4.20/21-pre. -aa if possible). I could 
also capture vmstat etc..

Thanks for your help.
-- 
Hari
harisri@bigpond.com


             reply	other threads:[~2003-03-13 13:42 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2003-03-13 13:51 Srihari Vijayaraghavan [this message]
2003-03-13 14:25 ` Torsten Foertsch
2003-03-13 14:28 ` OFFTOPIC: " uaca

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=200303140051.19453.harisri@bigpond.com \
    --to=harisri@bigpond.com \
    --cc=linux-kernel@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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®