Guys,
I wanted to ask the forum one question why DSCP AF-41 marking is used for Video. I know Cisco designed and recommedded it in that way, but is there any other technical reason behind it?
Cheers,
Sugata
已解决! 转到解答。
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
What you are referring by old equipment? Old equipment & old operating system? Is it something you have experienced or you are anticipating?
By old equipment and operating software, I'm referring to devices that predated DSCP. Don't see too much of that today, but when DSCP was first suggested most devices, if they supported the ToS byte, only "knew" of IP Precedence. So, DSCP was set up to be somewhat backward compatible. I.e. if a DSCP AF4y (or CS4) marked packet hit a network device still using IP Precedence rules, AF4y (or CS4) would likely get better treatment then default or lower DSCP class values.
And yes, do have experience using equipment that only supported IP Precedence. (Some of it, though, later supported DSCP with OS upgrades.)
An example of this, WFQ (not CBWFQ, except for FQ in class-default pre-HQF) on a router interface (i.e. fair-queue) actually often uses IP Precedence for adjusting scheduling priorities; higher is better (NB: Cisco has two WFQ weighting variants, that I'm aware of.)
best practices Qos implementation:
Generally, Cisco design guides recommend AF41 (DSCP value 100010) for video. There is no advantage if you treat the audio portion of the video streams better than the video packets in an IP video conference application. Therefore, use AF41 as the DSCP value for both voice and video media in a video conference.
source:http://www.cisco.com/en/US/tech/tk543/tk757/technologies_tech_note09186a0080094968.shtml
=============================
Please remember to rate useful posts, by clicking on the stars below.
=============================
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
Technically, any DSCP marking can be used for anything. By RFC standards, how to use DSCP markings are recommendations.
In RFCs, AF classes are used for some level of controlled service and AFxy markings might be subject to drops with drop probability possibly increasing AFx1, AFx2 and AFx3. (In other words, by recommendations, AF41 might obtain the best treatment. Only DSCP EF and network traffic [CS 6 and 7] usually obtain better.)
In the older IP Precedence standard, higher values are "more important". So as real-time video is considered "more important" (for providing it the bandwidth it needs to function correctly) it uses the highest AF class, 4, which corresponds to IP Precedence class 4. This so (hopefully) old equipment will treat such marked traffic "better". (BTW, L2 CoS and MPLS Exp are like IP Precedence too, so "4" might also be used for them to "better" service.)
PS:
Also BTW, there are a couple of minor differences between RFC DSCP usage recommendations and Cisco's recommendations.
Hi Joseph quick question on your last part of the answer:
"
In the older IP Precedence standard, higher values are "more important". So as real-time video is considered "more important" (for providing it the bandwidth it needs to function correctly) it uses the highest AF class, 4, which corresponds to IP Precedence class 4. This so (hopefully) old equipment will treat such marked traffic "better". (BTW, L2 CoS and MPLS Exp are like IP Precedence too, so "4" might also be used for them to "better" service.) "
What you are referring by old equipment? Old equipment & old operating system? Is it something you have experienced or you are anticipating?
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
What you are referring by old equipment? Old equipment & old operating system? Is it something you have experienced or you are anticipating?
By old equipment and operating software, I'm referring to devices that predated DSCP. Don't see too much of that today, but when DSCP was first suggested most devices, if they supported the ToS byte, only "knew" of IP Precedence. So, DSCP was set up to be somewhat backward compatible. I.e. if a DSCP AF4y (or CS4) marked packet hit a network device still using IP Precedence rules, AF4y (or CS4) would likely get better treatment then default or lower DSCP class values.
And yes, do have experience using equipment that only supported IP Precedence. (Some of it, though, later supported DSCP with OS upgrades.)
An example of this, WFQ (not CBWFQ, except for FQ in class-default pre-HQF) on a router interface (i.e. fair-queue) actually often uses IP Precedence for adjusting scheduling priorities; higher is better (NB: Cisco has two WFQ weighting variants, that I'm aware of.)
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
How Cisco equipment recognized DSCP EF, and how it might treat it, depends on the QoS features of the platform, and platform's configuration, either by default or as manually configured. Generally, EF is used for real-time extremely delay sensitive traffic, so often QoS is configured to deschedule it first. Again though, depends on the platform's QoS configuration. EF, alone, without any supporting QoS configuration support is treated like any other traffic.