VL2: A Scalable and Flexible
Data Center Network
Albert Greenberg, James R. Hamilton, Navendu Jain,
Srikanth Kandula, Changhoon Kim, Parantap Lahiri,
David A. Maltz, Parveen Patel, Sudipta Sengupta
Microsoft Research
SIGCOMM Comput. Commum. Rev., Vol 39, No. 4. (2009) , pp.51-62
1
Outline
• Motivation
• Conventional Data Center Architecture
• Virtual Layer 2 (VL2)
– Design
– Advantage
• Evaluation
• Conclusion
2
Motivation
• The network is a bottleneck to Data Center
computation.
• Today’s data center network has several
issues:
– Tree architecture
– Congestion and computation hot spot
– Traffic Independence
– IP configuration Complexity
– Migration
– Tradeoff
– Reliability and utilization
3
Conventional Data Center
Architecture
4
Design of VL2
• Location-specific IP address(LAs)(public)
• For all switches and interfaces or external
server
• Application-specific IP addresses
(AAs)(private)
• For application servers
• VL2 Directory System
• Stores the mapping of AAs to LAs
• Access control
5
Scale-out Topologies
Aggr : Int = n:m
ToR : Aggr = 1:2
6
Scale-out Topologies
Benefit
• Risk balancing
– The failure of a Int. reduces the bisection
bandwidth by only 1/m
• Routing is extremely simple on this
topology
– Random path up and random path down.
7
Example
Randomly
select Int.
by ECMP
Create by
VL2 Agent
8
VL2 agent
– VL2 agent’s work flow
 Intercepts the ARP request for the
destination AA
 Converts ① to an unicast query to the VL2
directory system
 Intercepts packets from the host
 Encapsulates the packet with the LA
address from ②.
 Caches the mapping from AA to LA
addresses
9
Advantage of VL2
• Load Balance
• Randomizing to Cope with Volatility:
• Building on proven networking technology:
•
•
•
•
link-state routing,
equal-cost multi-path(ECMP) forwarding,
IP anycasting,
IP multicasting
• Simple Migration
– Static AAs
– Only need to update AAs & LAs mapping
• Eliminating the ARP and DHCP scaling
bottlenecks
10
Evaluation: testbed
• 80 servers
• 5 for directory system
• Intermediate switches*3
• 24 10Gbps Ethernet ports(3
for Aggr)
• Aggregation switches*3
• 24 10Gbps Ethernet ports(3
for Aggr, 3 for ToR)
• ToR*4
• 24 1Gbps ports
11
Experiment objective
• We test the following three objectives
– Uniform high capacity
– Fairness
– Performance isolation
12
VL2 Provides Uniform High
Capacity
• We create an all-to-all data shuffle traffic matrix
involving 75 servers.
– Each of 75 servers deliver 500MB data other 74 servers a shuffle of 2.7 TB from memory to memory.
– During the run, the sustained utilization of the core links in
the Clos network is about 86%. (more than 10x what the
network in our current data centers can achieve with the
same investment)
13
VL2 Provides Fairness
– In the experiment before, we observe the 3 aggregate
switch
– We use Jain’s fairness index[15]
[15 R.Jain. The Art of Computer Systems Performance analysis, techniques for Experimental Design, Measurement, Simulation,
and Modeling. John Wiley & Sons, INC, 1991.
14
VL2 Provides
Performance Isolation
15
Conclusion
• VL2 provides
• The simpler abstraction that all servers
assigned to them are plugged into a single
layer 2 switch
• Hotspot free performance
• A simple design that can be realized today
• High utilization
• Achieves high TCP fairness.
16
Q&A
I HAVE QUESTIONS!!
17