Re: [leaf-user] Assigning ethernet interfaces to physical NIC's.

[Lyndsay Roger]

On Sun, 2003-04-06 at 10:52, Laleem wrote:
 Hello,
 
 I have some questions about the network cards and the ethx interfaces. How 
 do the ethernet interfaces get connected to the physical network cards?
 

The connection is made when the module is loaded into the kernel.
If you have two cards that use the same module then the module will scan
the pci bus and the first card that it detects will be allocated eth0
and the next eth1 and so on.
If you have two different cards using different modules then the order
the modules are loaded is important. The first module loaded will scan
for cards and start allocating from eth0 and the second or later modules
will scan for cards and start allocating from the next spare ethx
number.


 Also, is there a way for me to force the network card in the upper PCI slot 
 to take eth0 (and so on)?

If the two cards use the same module then I am not sure if you can
change the order(someone else may know) but if the cards are different
and use different modules then change the order of loading the modules
in /etc/modules

  The cards I have are PnP and apparently I cannot 
 hard configure the IRQ or I/O range.

What type of cards are they?


-- 
Lyndsay Roger
Dunedin, New Zealand




---
This SF.net email is sponsored by: ValueWeb: 
Dedicated Hosting for just $79/mo with 500 GB of bandwidth! 
No other company gives more support or power for your dedicated server
http://click.atdmt.com/AFF/go/sdnxxaff00300020aff/direct/01/

leaf-user mailing list: [EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/leaf-user
SR FAQ: http://leaf-project.org/pub/doc/docmanager/docid_1891.html

Re: [leaf-user] RE: Bering ip_conntrack_max on a 21Port, heavyloaded router

[Lyndsay Roger]

On Fri, 2003-01-17 at 05:50, Sandro Minola wrote:
  What does the value 32768 mean? Max. size of the conntrack table
  in bytes? Max. # of entries in the table?
 I just found out that this value means the number (#) of connections. And
 it's currently not set to 32768 (as I misleadingly wrote), it's set to
 32760.
 http://www.sns.ias.edu/~jns/security/iptables/iptables_conntrack.html says
 that this is the default value for a 512MB RAM machine.
 So it seems that my router can handle 32760 simultaneous connections. Do you
 think this is enough for 200 clients and 20 servers?

At a guess, and only a guess, I don't think you will have problems. You
can increase the limit by -
echo 9 /proc/sys/net/ipv4/ip_conntrack_max

Change 9 to what ever you think you need by remember that each
connection uses about 350 bytes of memory so 9*350 is about 35MB but
with 512MB to play with I don't think it will be a problem :-))

If you have concerns then you may want to run a script file that
monitors the number of connections and reports to you by email.

Add a file with -
date -R /tmp/connections
cat /proc/net/ip_conntrack | wc -l /tmp/connections


and get cron to run this every 5 minutes or so and then each hour, 6
hours or what ever you want run the following by cron -
cat /tmp/connections | mail -s Connections report [EMAIL PROTECTED]
rm -f /tmp/connections



This should produce an email something like-
Fri Jan 17 11:23:18 NZDT 2003
 31
Fri Jan 17 11:23:42 NZDT 2003
 36

With this you can track the load during the day to see what your peak
connections are and also the time of day it happens.

Run it for a week or so to see how your system is going.

If you want to take it a step further then you could create a script
that only emails an alert or warning if the number of connections get
above a certain amount. 

-- 
Lyndsay Roger [EMAIL PROTECTED]



---
This SF.NET email is sponsored by: Thawte.com
Understand how to protect your customers personal information by implementing
SSL on your Apache Web Server. Click here to get our FREE Thawte Apache 
Guide: http://ads.sourceforge.net/cgi-bin/redirect.pl?thaw0029en

leaf-user mailing list: [EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/leaf-user
SR FAQ: http://leaf-project.org/pub/doc/docmanager/docid_1891.html