Limiting all outgoing traffic on eth0 is what we want here.

Clear any existing traffic shaping on eth0

 # tc qdisc del dev eth0 root

Add the root handle and give it a default traffic shaping algorithm "htb"

 # tc qdisc add dev eth0 root handle 1: htb default 26

Create a class of traffic, which will have a maximum output of 100kbit

 # tc class add dev eth0 parent 1: classid 1:1 htb rate 100kbit

Assign all traffic to address 91.195.201.27 to this class

 # tc filter add dev eth0 parent 1: protocol ip prio 0 \
   u32 match ip dst 91.195.201.27 flowid 1:1

To be tested: add the ingress queue:

 # tc qdisc add dev eth0 handle FFFF: ingress

Apply the filter straight away:

 # tc filter add dev eth0 parent ffff: protocol ip prio 2 \
   u32 match ip src 91.195.201.27 police rate 100kbit \
   burst 10kb drop flowid ffff:

Instead of the above source and destination address, it's entirely possible to shape all traffic with 0.0.0.0/0.

To show the status:

 # tc qdisc show dev eth0

Clear it if you're not happy:

 # tc qdisc del dev eth0 ingress

Bridging

TODO not finished yet

Now that we know how to shape, let's test it:

 # Create a bridge
 sudo brctl addbr br0
 sudo brctl stp br0 off

 # Make dummy interfaces
 sudo rmmod dummy
 sudo modprobe dummy numdummies=10
 sudo ifconfig dummy0 hw ether de:ad:be:ef:00:00
 sudo ifconfig dummy0 up
 sudo ifconfig dummy1 hw ether de:ad:be:ef:00:00
 sudo ifconfig dummy1 up
 sudo brctl addif br0 dummy0
 sudo brctl addif br0 dummy1