[mrtg] Bandwidth per user session - VPN 3000 & ASA

Steve Shipway s.shipway at auckland.ac.nz
Thu Dec 17 01:40:26 CET 2009

If you're getting a graph that jumps between maxbytes and zero, then possibly the value being retrieved/calculated is >maxbytes and so is not being recorded.  To test this, set AbsMax[] to something much bigger (eg 10xMaxbytes) and see if the line appears.  Of course this doesnt explain why it is so big,..


From: mrtg-bounces at lists.oetiker.ch [mrtg-bounces at lists.oetiker.ch] On Behalf Of Martin77 [mrtg at mg77.net]
Sent: Thursday, 17 December 2009 1:35 p.m.
To: mrtg at lists.oetiker.ch
Subject: [mrtg]  Bandwidth per user session - VPN 3000 & ASA

MRTG Guru's,

I've been looking at the conf files too long and can't figure out why this
works on one device, but not the other.

Trying to graph bandwidth utilization of a single user session connected to
a VPN gateway. Don't care about any particular users or top users or
anything, just a simple 'current interface utilization / current number
users = bandwidth per user' type expression.

Running MRTG 2.16.2.

These conf snippets running against a VPN 3000 work perfectly:
Target[brm-vpn-srh-1.bw_in]: 1:public at brm-vpn-srh-1:
Target[brm-vpn-srh-1.bw_out]: 2:public at brm-vpn-srh-1:
Target[brm-vpn-srh-1.bw_per_user]: 1:public at brm-vpn-srh-1 /
. at brm-vpn-srh-1

MaxBytes[brm-vpn-srh-1.bw_per_user]: 12500000
Options[brm-vpn-srh-1.bw_per_user]: bits

Graph pattern follows i/f utilization graph, with numbers scaled down to
appropriate value based on how many people logged in at that time.

On an ASA, I have these conf snippets:
. at sca-vpn-srh-1
Target[sca-vpn-srh-1.bw_in]: #inside:public at sca-vpn-srh-1-mgmt:::::2
Target[sca-vpn-srh-1.bw_out]: #outside:public at sca-vpn-srh-1-mgmt:::::2
Target[sca-vpn-srh-1.bw_per_user]: #outside:public at sca-vpn-srh-1-mgmt:::::2
/ . at sca-vpn-srh-1-mgmt:::::2

MaxBytes[sca-vpn-srh-1.bw_per_user]: 125000000
Options[sca-vpn-srh-1.bw_per_user]: bits

The result here is that the graph jumps up and down between almost nothing,
the 125M max set in MaxBytes.

I think what it's doing in the 2nd case is dividing the actual value of the
byte counter by user count.
Using indexes & version 1 like on the 3k makes no difference.

Can anyone see what I'm doing wrong to get different expression eval
behavior between these 2?
Any way to make it use the utilization of an interface in the expression,
instead of the raw counter value?

Thanks & Regards,
View this message in context: http://n2.nabble.com/Bandwidth-per-user-session-VPN-3000-ASA-tp4178963p4178963.html
Sent from the MRTG Users Mailinglist mailing list archive at Nabble.com.

mrtg mailing list
mrtg at lists.oetiker.ch

More information about the mrtg mailing list