07-31-2015 09:24 AM - edited 03-05-2019 01:59 AM
Good Idea or bad Idea?
I have a router with a HQF QoS policy configured, within the outbound policy the default class has IP precedence WRED configured. IPp 0 has many (increasing) drops for random and tail drops. The throughput used is far lower than the configured policy, I think this is microburst traffic which we don’t see on our NMS.
Whilst reading this Q&A http://www.cisco.com/c/en/us/products/collateral/routers/asr-1000-series-aggregation-services-routers/q-and-a-c67-731655.html I read that if you see drops but the throughput I "low" its probably microbursts and that you can try and double the queue-limit and up the WRED thresholds.
Has anyone else done this and resolved this type of issue?
Thank you
Solved! Go to Solution.
07-31-2015 10:18 AM
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've often adjusted queue-limit as part of my QoS. But then, I consider drop management an important, but often overlooked aspect of QoS.
Often increasing queue limits will decrease drops, but there are considerations beyond additional queuing latency (as mentioned in your referenced link).
In general, I recommend non-QoS experts avoid using WRED. At first glance it seems simple, but getting WRED to be an effective QoS component can be difficult.
07-31-2015 10:18 AM
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've often adjusted queue-limit as part of my QoS. But then, I consider drop management an important, but often overlooked aspect of QoS.
Often increasing queue limits will decrease drops, but there are considerations beyond additional queuing latency (as mentioned in your referenced link).
In general, I recommend non-QoS experts avoid using WRED. At first glance it seems simple, but getting WRED to be an effective QoS component can be difficult.
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide