The proliferation of cloud computing, virtualization and IoT (Internet of Things) drives the rapid growth of traffic between devices in the data center. In response, there comes a new tendency for network topology design standard—leaf-spine architecture. Which creates a fast, predictable, scalable, and efficient communication architecture for data centers. Most network professionals choose this leaf-spine topology to satisfy the current traffic demands while to get prepared for future capacity growth. This article offers some rudiments of leaf-spine architecture, illustrating how to optimize and scale on leaf-spine topology with FS.COM switches.
The leaf-spine design consists of only two layers, the leaf layer and spine layer. The leaf layer involves access switches that connect to endpoints like servers, storage devices, firewalls, load balancers, and edge routers. The spine layer is made up of switches that perform routing, working as the backbone of the network. It functions to provide inter-connectivity between the leafs, but the network endpoints do not connect to the spines. In leaf-spine architecture, every leaf switch is interconnected with each and every spine switch, and the spines are not interconnected with each other. With this design, all devices are exactly the same number of segments away. In addition, they contain a predictable yet consistent amount of delay or latency for traveling information.
The accelerating traffic between devices in the data center is referred to as “east-west traffic”—traffic going back and forth between servers. However, the traditional three tiered topology is typically designed for north-south traffic (traffic that enters and exits the data center). If running massive east-west traffic through this conventional architecture, devices connected to the same switch port may contend for bandwidth, resulting in poor response time obtained by end-users. Moreover, if hosts on one access switch need to quickly communicate with hosts on another access switch, uplinks between the access layer and aggregation can be a point of congestion. The three-tiered topology may make it worse.
Besides, leaf-spine topology also obtains the following key benefits:
- There are multiple redundant paths available for traffic between any pair of leafs, since each leaf is connected to every spine. Link failures cause other paths in the network to be used.
- The leaf-spine topology provides a basis for a scale-out architecture. New leafs can be added to the network without affecting the provisioned east-west capacity for the existing infrastructure.
- New spines and new uplink ports on the leafs can be provisioned to increase the capacity of the leaf-spine fabric.
- The role of each tier in the network is well defined, providing modularity in the networking functions and reducing architectural and deployment complexities.
- The leaf-spine topology provides granular control over subscription ratios for traffic flowing within a rack, between racks, and outside the leaf-spine topology.
Leaf-spine network topology gains in much popularity in today’s data centers. FS.COM has developed and launched a series of data center open networking switches to support this new architecture, including S5850-48S2Q4C, S5850-48S6Q, S8050-20Q4C, etc.
Here we will use a simple example. Give that we want to build a data center fabric with a primary goal of having at least 900 10G servers in one fabric with 3:1 oversubscription. In this case, we will take S5850-48S6Q and S8050-20Q4C as an example to show how to scale on leaf-spine architecture. S8050-20Q4C is a high performance 40G/100G Ethernet switch for data center Spine/Leaf switches with highlighted network visibility features, which supports L2/L3/Data Center/Metro features. It has 20 ports of 40G and 4 ports of 100G. S5850-48S6Q Ethernet switch is to meet next generation Metro, Data Center and Enterprise network requirements, offering 48 ports of 10G and 6 ports of 40G. The details of these two switch modules are presented as following.
In spine-leaf network architecture for 40G application, the connections between the spine switches and leaf switches are 40G, while connections between the leaf switches and servers are usually 1/10G. Thus the 40G QSFP+ ports of S5850-48S6Q switch can be used to connect the spine switch S8050-20Q4C, and the 10G SFP+ ports of S5850-48S6Q are suggested to connect servers and routers. Every leaf switch is connected to every spine. Therefore, the number of connections used for uplinks from each leaf determines the number of spine switches we can have (4 ports here for four spine switches). And the number of ports on each spine switch determines the number of leaf switches we can have (20 leaf switches here). So in building this leaf-spine architecture, the maximum amount of 10G servers is 960 at 3:1 oversubscription.
When cabling for leaf-spine architecture with FS.COM S5850-48S6Q and S8050-20Q4C switches, patch cables, optical transceivers or direct attach cables (DACs) are needed. The following are some related accessories for completing the whole connection in 40G leaf-spine network.
|40GBASE-CSR4 QSFP+||Generic Compatible 40GBASE-CSR4 QSFP+ 850nm 400m MTP/MPO Transceiver for MMF, MTP/MPO|
|40GBASE-SR4 QSFP+||Generic Compatible 40GBASE-SR4 QSFP+ 850nm 150m MTP/MPO Transceiver for MMF, MTP/MPO|
|40GBASE-LR4L QSFP+||Generic Compatible 40GBASE-LR4L QSFP+ 1310nm 2km LC Transceiver for SMF, LC duplex|
|40GBASE-PLRL4 QSFP+||Generic Compatible 40GBASE-PLRL4 QSFP+ 1310nm 1.4km MTP/MPO Transceiver for SMF, MTP/MPO|
|40GBASE-LR4 and OTU3 QSFP+||Generic Compatible 40GBASE-LR4 and OTU3 QSFP+ 1310nm 10km LC Transceiver for SMF, LC duplex|
|40GBASE-PLR4 QSFP+||Generic Compatible 40GBASE-PLR4 QSFP+ 1310nm 10km MTP/MPO Transceiver for SMF, MTP/MPO|
|40GBASE-ER4 and OTU3 QSFP+||Generic Compatible 40GBASE-ER4 and OTU3 QSFP+ 1310nm 40km LC Transceiver for SMF, LC duplex|
|40GBASE-LX4 QSFP+||Generic Compatible 40GBASE-LX4 QSFP+ 1310nm 2km LC Transceiver for SMF&MMF, LC duplex|
|40G QSFP+ Passive DAC||Generic Compatible 40G QSFP+ Passive Direct Attach Copper Cable|
|40G QSFP+ Active DAC||Generic Compatible 40G QSFP+ Active Direct Attach Copper Cable|
|40G QSFP+ to 4 SFP+ DAC||Generic Compatible 40G QSFP+ to 4x10G SFP+ Passive Direct Attach Copper Breakout Cable|
|40G QSFP+ AOC||Generic Compatible 40G QSFP+ Active Optical Cable|
|40G QSFP+ to 4 SFP+ AOC||Generic Compatible 40G QSFP+ to 4x10G SFP+ Breakout Active Optical Cable|
|10GBASE-SR SFP+||Generic Compatible 10GBASE-SR SFP+ 850nm 300m DOM IND Transceiver, MMF, LC duplex|
|10GBASE-LRM SFP+||Generic Compatible 10GBASE-LRM SFP+ 1310nm 220m DOM Transceiver, MMF, LC duplex|
|10GBASE-LR SFP+||Generic Compatible 10GBASE-LR SFP+ 1310nm 10km DOM Transceiver, SMF, LC duplex|
|10GBASE-ER SFP+||Generic Compatible 10GBASE-ER SFP+ 1550nm 40km DOM Transceiver, SMF, LC duplex|
|10GBASE-SR SFP+||Generic Compatible 10GBASE-SR SFP+ 850nm 300m DOM Transceiver, MMF, LC duplex|
|10GBASE-LR SFP+||Generic Compatible 10GBASE-LR SFP+ 1310nm 10km DOM IND Transceiver, SMF, LC duplex|
|10GBASE-ZR SFP+||Generic Compatible 10GBASE-ZR SFP+ 1550nm 80km DOM Transceiver, SMF, LC duplex|
|10G SFP+ Passive DAC||Generic Compatible 10G SFP+ Passive Direct Attach Copper Twinax Cable 30AWG|
|10G SFP+ Active DAC||Generic Compatible 10G SFP+ Active Direct Attach Copper Twinax Cable 30AWG|
|10G SFP+ AOC||Generic Compatible 10G SFP+ Active Optical Cable|
Leaf-spine network architecture shines out in data centers since it improves the total available bandwidth, simplifies network configuration and facilitates IT department’s management. Meanwhile, leaf-spine design also greatly enhances network stability and flexibility. FS.COM provide high performance switches that support leaf-spine architecture, as well as cabling options for accomplishing the 40G leaf-spine connection. For more details, please visit www.fs.com or directly contact us via firstname.lastname@example.org.
Copyright © 2002-2018. All Rights Reserved.