aboot the threshold value in the track

[langjunqing]

Hi,

I have google and asked many ppl about what does the "threshold" refer in the track command, but i still cannot understand it.Would u give me some explain by a this example? Thx very much.

for example:

        track 10 ip route 10.2.21.128 255.255.255.0 metric threshold
        threshold metric up 63 down 64

        standby 10 priority 200
        standby 10 track 1 decrement 60

63 and 64 is up and down value of the threshold. but what does this mean? between 63 and 64?

And what is this "threshold" refer, metric? ttl? if metric, why we can use it in standby,u see,the 200 is a "priority" not a  "metric".

Any help is appreciated this problem has confused me quite a long time

[Peter Paluch]

Hello,

The threshold values define when a tracked object shall be deemed usable or unusable.

Let's take a more visible example:

track 10 ip route 10.2.21.128/25 metric threshold                   
 threshold metric up 20 down 50

If the metric of the route is 20 or less, the entire tracking object 10 is in the valid state. Now, if the metric of the route starts to increase, the tracking object 10 will be in the valid state until the metric grows to the value 50. If the metric becomes 50 or more, the track object 10 will flip to the invalid state.

Assume now that the route metric will again start to decrease. It will decrease below 50 but the object will still remain in the invalid state. It will decrease further, and when it reaches 20 or less, the tracking object will once again become active.

Thus, the range between the up and down threshold creates a hysteresis field that may help to dampen unnecessary state changes and provide a certain degree of reserve in the tracked value once the object flips its state again.

Best regards,

Peter

[skarthic]

Excellent explanation..You have explained a good use of this(dampening).

[dmorrow486]

Great explantion! thank you.

It is also worth noting how the metric values are derived so that you have an upper and lower range to work with when dealing with different routing protocols.

The"threshold metric up down " (values are between 0 - 255)

Default up threshold value = 254

Default down threshold value = 255

So let's say you want to configure tracking for an EIGRP route. First you want to know what the scaled value is for EIGRP. Simply issue the following command on you Cisco router.

DSW1#show track resolution

Route type     Metric Resolution

static         10

EIGRP          2560

OSPF           1

ISIS           10

So if you have an EIGRP Route with Metric 284160 as per route show below. Therefore 284160 / 2560 = 111

You now have a scaled value which you might want to use as the up value.

D EX    10.1.1.8/30 [170/284160] via 10.1.4.5, 00:50:39, FastEthernet0/0

To set the down value follow what Peter has described as your acceptable range. Lets say you set the down metric to 112 then you have a very narrow hysteresis. The EIGRP metric for your down vlaue is 2560 x 112 = 286720

Regards,

Dave

[jlindsey76]

Wow, this is exactly what I was looking for.  Thanks for the command and thanks for the explanation.  :-)

[Dinesh Moudgil]

What I think  is that the metric used here is the one associated with the route.
And threshold refers to the metric concerned with the route itself.

My question is , how is this concept being related to HSRP here.
Please throw some light on this topic..

[HMakhlouf]

1.      track 10 ip route 10.2.21.128 255.255.255.0 metric threshold
2.        threshold metric up 63 down 64

3.       standby 10 priority 200
4.       standby 10 track 1 decrement 60

Line one configures an object tracking the mensioned route, tracking its metric to be precise.

Line two gives the track command the limits, threshold, according to which it sets the value of the object to up or down.

Line three enables HSRP and sets its priority to 200, remember the default value is 100 and the router with the higher priority becomes active, only if it is configured witht the 'preempt' keyword or there is no active router available yet.

Line Four, here is the answer to your question, tells the router according to the value returning from the tracking object, change the priority value of the HSRP instance 10.

ex: If the object is down  >> HSRP priority = 200( configured priority) - 60(configured decrement in line 4) = 140

if the object is up again >> HSRP priority = 200 again.

Please note:

I used the four lines provided in the question.However, line 1 and line 4 have to have to the same track number, either both 'track 1' or 'track 10'.

You can also have more than one object tracked by your HSRP instance.

Makh

[SUFYAN]

thanks for the information. my question is that we can accomplish the same result as below

1. track 10 ip route 10.2.21.128 255.255.255.0 reachability
3. standby 10 priority 200
4. standby 10 track 10 decrement 60

then what is the main reason of using threshold metric option ?

[ldmccalla]

The reachability keyword just means that the exact route exists in your route table.

suppose you have 2 HSRP routers and both have dynamic uplinks o the internet and to each other. HSRP serves to protect your LAN from a Default GW outage but suppose Router A loses its BGP connectivity yet is still functional.  The idea is to set the metric threshold such that if the hop count changes and the router detects that it has a poor route to the internet, it should decrease its priority.

if you used the reachability keyword then the GW would still be active even though your routing would be sub-optimal. ie Router A would be your default GW and it would forward all traffic to router B to reach the internet.

Leon