I have 300 stores and I'm looking for a global QOS config that would shape/policy the bandwith based on a % of the detected ISP link speed

[netadmincsm]

Hi,
We have 300 stores with mostly the same router config.
The router in each store is a cisco 891 router capable of doing QOS.
Each of our store is connected to internet through a local Internet Service Provider (ISP).  They all have different ISP.
Each store has different Internet plan with different speed, some are DSL, other are using cable modem.

The connection between the router and the modem is 100Mbps. But the real speed after the modem will be different depending of the Internet plan the store is using. Exemple : 20Mbps download / 2Mbps upload.

I'm looking for a standard QOS config that would be put in each routeur and that would shape/policy the bandwith based on a % of the detected ISP link speed.  The goal here is to prevent a single desktop computer of the store to take all available bandwith of the ISP link because of youtube , cloud photo upload, etc.

Because the wan of the router is in the modem at 100Mbps, I cannot use this bandwith indication for the % calculation. The ISP link speed (after the modem)  is much lower than that.  

In other words, I dont want to set specific qos value or bandwith manual value in each of my 300 router. Especially because the ISP speed is not the same in each site, it's a big job and it's changing all the time.  Stores owner boost their link, etc.

So, Is there a way within the router to detect the Wan real max speed and to assign dynamically % of the deduced value and assign it to different policy-map rules, etc.

regards
Jonathan D

[Matt Rudkowski]

I believe what you're looking for is what's referred to as Cisco Performance Routing.

See here for a good overview:

http://docwiki.cisco.com/wiki/PfR:Technology_Overview#Measuring_Performance_and_Bandwidth

I'm not sure if it's supported on the 891 router's, so hopefully someone from Cisco will chime in on this.

[Joseph W. Doherty · Accepted solution]

Disclaimer

The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.

Liability Disclaimer

In no event shall Author be liable for any damages whatsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.

Posting

I'm unaware of any (one) "built-in" feature that will do what you're desire.  (BTW, I don't believe PfR, as mentioned by Matt, will do what you desire either, but I've haven't used it in several years.)

I was looking to address a somewhat similar problem several months ago, and believe I could construct dynamic QoS shaping using an embedded device script working with IP SLA monitoring.  However, the "problem" site (using a cable modem) closed, so I didn't pursue it further.

An alternative approach would be to use as few policy models as possible.  For example if there are sites with 2, 3 or even 5 Mbps uploads, you might use a 2 Mbps shaper for each.  Of course, sites with more available bandwidth won't obtain the full advantage of it, but your maintenance support will be easier.  Conversely, especially for cable modems, running under their nominal speed might provide much more predictable performance, as HFCs are generally much more variable in their actual available bandwidth, moment to moment.