Backward Congestion Notification Version 2.0

1
Backward Congestion Notification Version 2.0

• Davide Bergamasco (davide_at_cisco.com)
• Rong Pan (ropan_at_cisco.com)
• Cisco Systems, Inc.
• IEEE 802.1 Interim Meeting
• Garden Grove, CA (USA)
• September 22, 2005

2
Credits

• Valentina Alaria (Cisco)
• Andrea Baldini (Cisco)
• Flavio Bonomi (Cisco)
• Manoj K. Wadekar (Intel)

3
BCN v2.0

• Desire from Mick to see an analytical studyof
  BCN stability
• BCN v2.0 improvements
• Linear control loop allows analysis of stability
• Simplified detection mechanism
• Reduced signaling rate
• Original BCN framework remains the same

4
BCN Background
5
Detection Signaling
6
Reaction
7
Suggested BCN Message Format
0 15
31 ------------
--------------------

DA SA of sampled
frame ----------------

------------
---- SA MAC Address of CP

------------
--------------------
IEEE 802.1Q Tag or S-Tag
------------
--------------------
EtherType BCN Version
Reserved -----------
---------------------


CPID

-----------
---------------------
Qoff
Qdelta ----------
----------------------
Timestamp
----------
------

First N
bytes of sampled frame starting from DA


-------------------------
-------
FCS
-------------------------
-------
8
Suggested RLT Tag Format
0 3 7 15
31 -----------
---------------------

DA of rate-limited
frame ----------------

-----------
----- SA of rate-limited frame

-----------
---------------------
IEEE 802.1Q Tag or S-Tag of
rate-limited frame -----------
---------------------
EtherType RLT Version
Reserved -----------
---------------------


CPID

-----------
---------------------
Timestamp EtherType of
rate limited frame --------------
------------------

Payload of rate-limited
frame

------------------------
--------
FCS
------------------------
--------
9
Simulation Environment (1)
TCP Bulk
UDP On/Off
Congestion
10
Simulation Environment (2)

• Short Range, High Speed DC Network
  
• Link Capacity 10 Gbps
• Switch latency 1 ?s
• Link Length 100 m (0.5 ? s propagation delay)
• Control loop
• Delay 3 ?s
• Parameters
• W 2
• Gi 4
• Gd 1/64
• Ru 8 Mbps
• Workload
• ST1-ST4 10 parallel TCP connections transferring
  1 MB each continuously
• SU1-SU4 64 KB bursts of UDP traffic starting at
  t 10 ms

11
BCNv1.0
12
BCNv2.0
Faster Transient Response
Higher Stability _at_ Steady State
13
Simulation Environment (3)

• Long Range, High Speed DC Network
  
• Link Capacity 10 Gbps
• Switch latency 1 ?s
• Link Length 20000 m (100 ? s propagation
  delay)
• Control loop
• Delay 200 ?s
• Parameters
• W 2
• Gi 4
• Gd 1/64
• Ru 8 Mbps
• Workload
• ST1-ST4 10 parallel TCP connections transferring
  1 MB each continuously
• SU1-SU4 64 KB bursts of UDP traffic starting at
  t 10 ms

14
BCNv1.0
15
BCNv2.0
Much higher stability _at_ steady state with larger
loop delays
16
Summary

• BCN v2 has a number of advantages
• Can be studied analytically
• Better protection of TCP flows in mixed TCP and
  UDP traffic scenarios
• Detection algorithm independent of Switch
  implementation
• Better Performance
• Lower signaling frequency (from 10 to 1)
• Better stability
• Increased tolerance to loop delays
• and one disadvantage
• Slower convergence to fairness

17
A Control-Theoretic Approach to BCNDesign and
Analysis
18
Notation
N Number of Flows C Link Capacity ? Round
Trip Delay w Weight of the Derivitive Pm
Sampling Probability Gi Additive Increase
Gain Gd Multiplicative Decrease Gain
19
Block Diagram of BCN Congestion Control
C

q
R

?R
_


Gd
_
Time Delay

Pm
Gi
20
Non-linear Differential Equations
Link Control
Source Control
If Fb(t-?) gt 0
If Fb(t-?) lt 0
21
Linearization Around Operating Point

• Using feedback control to analyze local stability
• Operating point
• R C/N
• q qeq q 0
• Linearization
• Difficulty depending on sgn(Fb(t-d)), the system
  responses are different
• Luckily, a piecewise-linear function
• Details are in the appendix

22
Block Diagram of BCN Feedback Control
?R

?q

lose 90o margin
Multiplicative Decrease
_
?Fb
Additive Increase

23
The Effect Of Zero From Time Domains Eyes
R
q
24
Choosing Parameters an example

• Network conditions (10G link)
• N 50
• ? 200us
• Choose parameters such that the feedback loop is
  stable with a 35o margin
• w 4
• Gi 2Mbps
• Gd 1/128
• Pm 0.01

25
Stability Result
lost 90o margin
26
Simulation Result Shows A Stable System for N
50 Delay 200us
27
Simulation Result Shows System is stable, but on
the verge of oscillation N 50, Delay 1ms
28
Change W 4 -gt 1
29
Indeed System is stable, but on the verge of
oscillation even for N 50, Delay 200us when w
1.0
30
Requests to 802.1

• Start a Task Force on Congestion Management
• Use BCN as a Baseline Proposal

31
Appendix
32
Linearizing
33
Linearizing Additive Increase Function
34
Linearizing Additive Increase Function
35
Linearizing Multiplicative Decrease Function
36
Linearizing Multiplicative Decrease Function
37
Issue 1 Non-linearity
Q

• ISSUE Overshoots and undershoots accumulate over
  time
• SOLUTION Signal only when
• Q gt Qeq dQ/dt gt 0
• Q lt Qeq dQ/dt lt 0
• Easy to implement in hardware just an Up/Down
  counter
• Increment _at_ every enqueue
• Decrement _at_ every dequeue
• Reduces signaling rate by 50!!

Stop Generation of BCN Messages
-
-
-
-




Qeq
t
38
Issue 2 Specific Detection Mechanism
39
39
39
39