1. Congestion control in data centers
Sruta Keerti Kasula

2. Data center structure
SRU-Server Request Unit

3. Incast Caused by Partition/Aggregate. TCP timeout Worker 1 Aggregator
RTOmin = 300 ms
Worker 4
TCP timeout

4. Why develop XCP?
TCP is 25 years old, designed for low bandwidth delay product.
The higher the Bandwidth-Delay product, the less efficient TCP’s congestion control algorithms such as AIMD.
Networks with high delays makes TCP react slower to packet drops. Additionally, bandwidth is reacquired more slowly in such networks.

5. XCP protocol
Application
TCP
XCP IP
Link
The TCP/XCP/IP stack
XCP introduces a 20 byte header between IP and TCP that carries information about the sender’s desired bandwidth and information about what the routers allow.
XCP requires all routers and hosts in the network to use the XCP protocol in order to work as intended.

6. XCP’s solution
XCP allows the routers in the network to continuously do the adjustments by changing the contents of the packets (XCP header) transferred between the sender and receiver.
XCP provides a theoretically analysis, yet effective approach to congestion control.
Router
Router

7. An XCP network (simplified)
Network RTT: 100 ms
Router’s capacity: B/s (available B/s)
Sender’s capacity: B/s (available B/s)
Sender’s current throughput: 0 B/s (or 0 B/ms) 1 6 20
100 1 6 20
100 1 2 6 20
100
RTT
Current Throughput
Delta Throughput
Feedback
Sender
Router
Receiver

8. An XCP network (simplified)
Network RTT: 100 ms
Router’s capacity: B/s (0 B/s is available)
Sender’s capacity: B/s ( B/s is available)
Sender’s current throughput: B/s (or 200 B/ms) 1 6 20
100
200 1 6 20
100
200 1 2 6 20
100
RTT
Current Throughput
Delta Throughput
Feedback
Sender
Router
Receiver

9. DCTCP ECN = Explicit Congestion Notification ECN Mark (1 bit) Sender 1
Receiver
Sender 2

10. Data Center TCP Algorithm
Don’t
Mark K
Switch side:
Mark packets when
Queue Length > K.
Mark
DCTCP satisfies all our requirements for Data Center packet transport.
Handles bursts well
Keeps queuing delays low
Achieves high throughput
Features:
Very simple change to TCP and a single switch parameter K.
Based on ECN mechanisms already available in commodity switch.

11. ICTCP Observations
The available bandwidth at the receiver is used as a quota to increase the receiver window size on all incoming connections.
Congestion control is performed for each flow independently in slotted RTT and this is also the control latency in the feedback loop.
A receive window based scheme is considered i.e., Ratio of difference of expected and measured throughput over expected is used to adjust the receiver window size.

12. ICTCP
It’s a window based congestion control algorithm at TCP receiver-side.
The experimental results demonstrated that ICTCP was effective to avoid congestion by achieving almost zero timeout for TCP Incast, and it provided high performance and fairness among competing flows

13. References folk.uio.no/paalh/students/PetterMosebekk.ppt
ast_in_Data_Center_Networks
4/2.pdf?id=7557
p-incast-congestion-control-for-tcp-in-data-center-sigcomm
Haitao Wu? et al., ICTCP: Incast Congestion Control for TCP in Data Center Networks, proceedings ACM CoNEXT 2010
V. Jacobson and M. Karels, Congestion Avoidance and Control, In Proc. ACM SIGCOMM '88. understand congestion problem