cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1864
Views
0
Helpful
4
Replies

Enterprise QoS - BW percent vs remaining

dlstokes
Level 1
Level 1

I sat in a great Enterprise QoS breakout session during CiscoLive this year regarding standards updates and recommendations for QoS policies facing SP-managed L3 VPN's.  During that session, I noticed that the recommended settings for all of the CBWFQ's was ...

bandwidth percent (x)

Traditionally, throughout our WAN Edge QoS policies on our CBWFQ's, we have always used ...

bandwidth remaining (x)

I was wondering what the significant difference was between 'remaining' and 'percent' and whether there are specific reasons why 'percen't is recommended.

TIA

Daniel

4 Replies 4

Richard Michael
Cisco Employee
Cisco Employee

Hello Daniel,

bandwidth percent {value} - Specifies bandwidth allocation as a percentage of the underlying link rate.

bandwidth remaining percent {value} - Specifies bandwidth allocation as a percentage of the bandwidth that has not been allocated to other classes.

Reference :


http://www.cisco.com/en/US/tech/tk543/tk757/technologies_tech_note09186a0080103eae.shtml#howisunusedbandwidthallocated

Thanks,

Ricky Micky

Peter Paluch
Cisco Employee
Cisco Employee

Hello Daniel,

In my personal opinion, the bandwidth percent is not about a best practice or a recommendation, rather simply about meeting your QoS requirements (for which this command may ultimately not be suitable).

Using the bandwidth percent, you have defined the ratios of guaranteed bandwidths for individual classes, however, without any absolute value for any class. The bandwidth remaining percent allows you first to assign an absolute bandwidth guarantee (bandwidth, priority) to a set of classes if you need it, and split the remaining capacity among other classes. WIth this approach, some of the classes have fixed and explicite bandwidth guarantees, as opposed to bandwidth percent command.

I do not think there is any general advantage to using any of these commands, and you should always use the command that creates the QoS policy according to your needs.

Best regards,

Peter

Marwan ALshawi
VIP Alumni
VIP Alumni

Hi Daniel

it just more reasonable to have your QoS guaranteed bandwidth to be % of the actual bandwidth but keep in mind that you need to shape the interface to reflect the actual available bandwidth to you from the other end over the WAN for example if you have Fa interface of 100M and the WAN link is 1M you need to shape the interface to 1M and under this shaping policy you aply the QoS policy with percentage otherwise the percentage will be based on the 100 not 1Mbps

about the bandwidth and renaming bandwidth

the first one is absolute bandwidth reservation while the second one is releive share of bandwidth reservation

for example if you have interface with 1M and you allocated 20% for LLQ and 10 bandwidth remaning to other class the other class bandwidth will 10% of ( 1M - 200kLLQ) which is 10% of 800K

while if you specify it as bandwidth % only it will be 10% of the 1M

please refer to the bellow link for more detail about how to configure HQoS and shape the interface using %

https://supportforums.cisco.com/docs/DOC-8373

HTH

if helpful Rate

Joseph W. Doherty
Hall of Fame
Hall of Fame

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

The difference, as others have already noted, bandwidth percent allocates against what it believes the total bandwidth to be, so bandwidth percent 50 would be 50 Mbps of a 100 Mbps Ethernet link (by default).  Bandwidth remaining 50 accounts for both what it believes the total bandwidth to be and what's already been otherwise allocated.  I believe this also accounts for class-default, which I further believe defaults to 25%.  So bandwidth remaining 50 would be (by default) .5 * .75 = 37.5 Mbps of a 100 Mbps Ethernet link.

Since both are relative, you can apply the same policy on different bandwidth links, e.g. E1 vs. T1.

Bandwidth remaining is a bit less likely to break if you make some changes to the policy, such as increasing max-reserved, increasing bandwidth for a defined class or even adding another defined bandwidth class.

The reason percent might be recommended, it's closer to an absolute bandwidth allocation.

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community:

Innovations in Cisco Full Stack Observability - A new webinar from Cisco