HSRP can be Super-Fast

An interesting thing I recently learned was that HSRP can be fine-tuned to be much much faster than the open standard VRRP. You can simply adjust the default timers. To match the speed of VRRP, you can change the hello timer to 1 second and the holddown timer to a value of 3. The default values for HSRP are hello=3 secs and holddown=10 secs.
To really ramp up the speed of the failover, you can define the timers in milliseconds! This will generate a ton of traffic between your devices but that's usually OK because most likely you will be using 100MBPS or even 1Gbps redundant links.
Here is the command to adjust the timers:
Switch(config-if)# standby timers [msec] [msec]

Interface tracking in HSRP

The coolest thing about HSRP is it's ability to track interfaces and decrement the switch's HSRP priority in response to interface state changes. With HSRP, the highest priority wins the active switch role. All other switches with interfaces in the same HSRP group are designated as standby switches. The default HSRP priority is 100 so if you have to manually change the priority to something higher if you want a particular switch to be the active switch in the group. It's important to remember to use the preempt feature also to ensure that the manually designated active switch returns to it's active role after an outage.

HSRP could be used to determine the best route out of the LAN by using interface tracking. If you have a switch with a bouncing WAN link then you probably don't want it to function as the active switch and therefore the default gateway for your LAN hosts.

Let's say that you have a couple layer 3 switches in your distribution layer running HSRP. Each switch may have redundant WAN links to your ISP. You can set up interface tracking on the WAN links so that each time one of the links drops, the HSRP priority of the switch decrements. Eventually, the priority is lower than one of the standby switches and that switch will take over the active role. This ensures that network traffic is using the most reliable switch to reach the outside world. This is just one example and there are a lot of uses for interface tracking.

The default decrement value is 10 meaning that each time the tracked interface(s) drops, the HSRP priority is decremented by 10. You can change this to any value you want. You can track multiple interfaces, like this:

interface ethernet0

ip address 10.1.1.1 255.255.255.0

standby ip 10.1.1.3

standby priority 110

standby track serial0

standby track serial1

This is a really nifty way to ensure you are using the best path for your traffic flow -- kinda like route dampening in BGP. The other alternative is to change the decrement value so that it immediately becomes a standby router and abdicates the active role. It's very flexible.

HSRP virtual MAC

A quick way to determine if you are running HSRP is to look at your arp cache from the command line of your windows or linux host. If you see the mac address 00-00-0c-07-ac-xx then you are running HSRP somewhere in your broadcast domain. This is a virtual MAC that is associated with a virtual IP address (VIP). It is likely that the VIP is serving as your default gateway.
We can breakdown the HSRP virtual MAC. The 00-00-0c portion is the Cisco vendor ID while the 07-ac section defines the HSRP protocol. The final 2 bits are reserved for the HSRP group number -- obviously expressed in hexidemical.

Redundancy Madness

I like redundancy solutions -- they're fun to configure. Conceptually they are simple. I think the complexity arises when they are used for load balancing. For the CCNP, Cisco press presents three redundancy technologies: HSRP, VRRP, and GLBP. Here's a breakdown of each:

HSRP (Hot Standby Routing Protocol)

-Cisco proprietary
-uses a default hello timer of 3 seconds with a hold timer of 10 seconds
-can load balance using multiple HSRP isntances or 'groups'

VRRP (Virtual Routing Redundancy Protocol)

-open standard
-faster timers than HSRP by default -- hello=1 second, hold timer = 3 seconds

GLBP ( Gateweay Load Balancing Protocol)

- Cisco proprietary also
-just like HSRP except it's designed for easier load balancing because it allows an active-active configuration

That's the bare basics of all three. In the following week I will have a series of posts that delve into the various complexities of these three redundancy protocols..

First Post

So I've decided to join the crowd and blog about my CCNP study experience. I hope to continue this blog as a CCIE blog within the next year. So far, I've passed the routing exam which I felt was very difficult due to the amount of material covered. Currently I am knee-deep in the switching exam material. I've covered up to advanced STP in the Cisco press book. Now I'm working on inter-VLAN routing and redundancy solutions. Hopefully as my studies progress, I will provide some interesting information to those interested in networking and the CCNP certification.