Traffic Graph Difference between Graph and textual IP bandwith
we had yesterday a bandwith problem with was first very strange because the traffic graphs show no much traffic on the inner firewall but on the outer one. After some research we found then that one telephone VLAN created much traffic but even when graphs show about 50 MBit traffic the IP based bandwith was nearly zero. :(
With further searching we found then that one employee was so brainfull to connect the phone twice in the network. Normally the switches should handle loop detection / broadcast storm but seems to failed here… I guess because SNOM phones have different MAC Addresses on NET/PC port.
So here the interesting question is why such can't be shown in WebGUI ... I luckily found my debian package nettop as jnettop agein in Debian and also FreeBSD/pfsense so I could discover the problem on console...
I take a look in the script how the bandwith is calculated. This script is called - for instance with this parameters:
: /usr/local/bin/rate -i lagg1_vlan85 -nlq 1 -Aba 20 -R -c 192.168.85.0/24
The interesting question is if it can be extend to show also global broadcast traffic like this:
: /usr/local/bin/rate -i lagg1_vlan85 -nlq 1 -Aba 20 -R -c 0.0.0.0/255.255.255.255
: /usr/local/bin/rate -i lagg1_vlan85 -nlq 1 -Aba 20 -R -c 255.255.255.255/0.0.0.0
the program itself seems not directly modified over 10 years:
: /usr/local/bin/rate -h
rate 0.9 - Mateusz 'mteg' Golicz firstname.lastname@example.org, 2003
usage: /usr/local/bin/rate [-h | -?]
/usr/local/bin/rate [mode select option] [-h | -?]
/usr/local/bin/rate -L <name>/usr/local/bin/rate [filtering/generic options] [mode select option] [mode options]