1

QoS - Color Aware and Color Blind mode.

  1. Is color - red, yellow and green is a kind of analogy for violate, exceed and conform traffic rate. If so, whey did they come for the colors here.

I mean there is no where in the packet or frames the color bits are stored. I think it is internal to the system. Am I correct? If not then where is it stored in the frame?

  1. What's the use of pre-color then ?

  2. What is color blind? Is it like either conform or violate? That is there is no such yellow.

Dual Rate - 3 color meter talks about -

2 buckets that's why two rate - one for CIR and another for PIR.

next 3 colors are conform - green, exceed - yellow and violate - red.

Why they have used colors here - just conform is enough and why it needs to be called with color green also. why exceed terminology is enough and why it needs to be called with yellow color? Why violate terminology is enough and why it needs to be called with red color?

Please explain, I am finding these colors and color blind things quite confusing.

  • Did any answer help you? If so, you should accept the answer so that the question doesn't keep popping up forever, looking for an answer. Alternatively, you can provide and accept your own answer. – Ron Maupin Dec 14 '19 at 20:59
2

Colors are just an analogy for humans.

QoS is realized with different traffic queues in a device and packets/frames are mapped to these queues by your QoS rules.

E.g. you have a high, a standard, and a low priority queue. You map VoIP to high, ICMP to low and keep the rest standard. You set the queue scheduling to 3:1, so in case of congestion 3 high packets are forwarded, but only 1 standard packet (or 3 standard packets for 1 low packet). Packets exceeding its queue capacity are dropped.

Additionally, you can police traffic using QoS rate limiting. E.g. you limit YouTube traffic to 1 Mbit/s per user or 10 Mbit/s for the whole network (if your device supports this kind of control). Excess rate traffic is dropped.

  • I think in differentiated service policy, each hop must distinguish packets from each other to map them on correct queues. I understand that ICMP, VOIP or other kinds of packets could be differentiated by the DSCPfield, for example. Anyway, policer introduces 3 "domains"(conformance, exceed and violate) which define each hop behavior toward packets.... and for a differentiated policy, there must be a way to distinguish the right domain for each received packets, ya? For example, how a color-aware router knows the received packet has exceeded the defined rate in a previous hop? – A.A Dec 12 '19 at 8:00
  • 1
    QoS mappings are on each hop individually. For a working policy across your network these need to match. The DSCP classes are independent from the used protocols but a hop may be configured to override the received DSCP values based on source/destination subnet, IP, L4 protocol or L4 port. The policing domains are used to configure the desired thresholds - you wouldn't want to have high-priority video streams suppress all other kinds of traffic. – Zac67 Dec 12 '19 at 10:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.