[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [cobalt-users] [RAQ3] & [RAQ4] Total traffic usage?
- Subject: Re: [cobalt-users] [RAQ3] & [RAQ4] Total traffic usage?
- From: Paul Warner <pwarner@xxxxxxxxxxxxxxxxxx>
- Date: Fri Apr 4 14:49:02 2003
- List-id: Mailing list for users to share thoughts on Sun Cobalt products. <cobalt-users.list.cobalt.com>
> We have a couple of customers who keep all their mail on the Cobalt
servers, rather than collecting it with a POP3 client, they simply retrieve
the headers and leave it on the server - several of their users have a lot
of mail, and check their mail from multiple PCs in multiple locations on a
continuous basis.
>
> Example:
>
> client A has 10 users, 2 of which are "super important" and can operate
out of each of their 3 offices anywhere in the state of Texas.
>
> So - in order to have their email in each location, they leave it on the
server, pulling headers down from each PC.
>
> Each of these "super important" users, has three PCs, pulling email
headers from 3 PCs, EVERY MINUTE OF EVERY DAY.
>
> They refuse to log-off, or turn off the PC, or even quit the mail program
when they leave a particular office - so the email checking is MASSIVELY
EXCESSIVE (in my, or any sane tech persons opinion I would hazard).
>
> anyway - they have disk space and a traffic quota - but the cobalts only
appear to provide stats on size of emails handled - how can I factor in all
this load and continual checking into their traffic quotas? Surely this is
adding up to a fairly significant amount to traffic - it has GOT to be?!?!
>
> They are prepared to pay, I just need to be able to calculate how much
they owe - and be able to prove it!
>
> Does anyone have any tools, scripts to generate total usage stats for a
virtual site? They have their own IP if that helps.
>
> thanks
>
> GHL
Greg-
You could use ipchains to pick off the traffic from the 3 IPs...see archives
for install and sources...I believe Bruce Timberlake made a pkg...
This creates the chain...not needed on each load of ipchains, just once to
create the chain.
/sbin/ipchains -N superguy1
This creates an accounting rule in the chain named superguy1
/sbin/ipchains -A superguy1 -i eth0
This makes a packet go through the superguy1 chain...one for each IP
/sbin/ipchains -A input -i eth0 --source xxx.xxx.xxx.xxx -j superguy1
To see results...
/sbin/ipchains -L superguy1 -n -v
The first two columns will show pkts and bytes. The -n prevents time wasted
resolving the IP to a name. You could embellish the rules with ports if you
want more data on the POP v. other traffic as in:
/sbin/ipchains -A superguy1 -i eth0 --destination-port 110
/sbin/ipchains -A superguy1 -i eth0 --destination-port ! 110
There is also an MRTG gizmo in the
/home/local/src/mrtg-2.9.27/contrib/ipchainacc that can be edited to turn
the collected data into pretty graphs that can be posted in a self-updating
web page...for you or the superguys
-- Paul