#2 sounds reasonable, esp as ntop's processing time is more dependent upon
the # of packets than their length (other than a few protocol-specific
decodes (http, ftp, etc.), we mostly deal with the packet header).

-----Burton


-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Craig
Humphrey
Sent: Monday, October 20, 2003 5:10 PM
To: '[EMAIL PROTECTED]'
Subject: RE: [Ntop] Dropped by the kernel


It's probably also worth checking out a couple of things:

1.  Turn off ntop and just use tcpdump on the interface.  What kind of stats
do you get?  Then turn of DNS resolution for tcpdump and run it again.  Same
or better?
2.  Check to see if you're generating large numbers of small packets.  I've
found that even a P3-1.4GHz 512Meg RAM, has trouble keeping up with large
numbers of small packets (possibly a combination of [hardware] platform,
NIC, NIC drivers and the linux kernel version).


Just my 2c.

Later'ish
Craig

> -----Original Message-----
> From: David Touitou [mailto:[EMAIL PROTECTED]
> Sent: Tuesday, October 21, 2003 10:02 AM
> To: [EMAIL PROTECTED]
> Subject: Re: [Ntop] Dropped by the kernel
>
>
> Burton M. Strauss III wrote:
>
> > Get a faster NIC/Processor.
>
> P3-733 (133 MHz FSB) and Intel 82559 Pro/100 Ethernet (with microcode
> loaded, aka less interrupt calls)) are not fast enough for 9Mbps
> bandwidth stats ?
>
> Wow...
>
> What kind of hardware should be used here ?
...
_______________________________________________
Ntop mailing list
[EMAIL PROTECTED]
http://listgateway.unipi.it/mailman/listinfo/ntop

_______________________________________________
Ntop mailing list
[EMAIL PROTECTED]
http://listgateway.unipi.it/mailman/listinfo/ntop

Reply via email to