In today’s data center environment, routing and switching architectures must evolve to support changing requirements. The advent of server farms, virtualization, and cloud computing structures is changing the optimal switching architecture.

This issue of the PTS Data Center Solutions e-newsletter focuses on the Modern Data Center’s traffic requirements and the impact on routing and switching and architecture.

As a recent new partner of Juniper, we are excited to be able to provide solutions which have evolved from carrier-class for use in the enterprise.

As always, I welcome your feedback on this issue of our newsletter. Please think of us for your next data center project — whether infrastructure, facility, or IT related.

 Pete Sacco

President & Founder

PTS Data Center Solutions, Inc.

Experts For Your Always Available Data Center

PTS Story: The Effects of East-West Traffic Flow on Routing & Switching Architectures in the Modern Data Center
The modern data center has undergone a series of changes that have significantly impacted business operations. Traditional data center traffic (e.g. web access at the user level) traverses the data center north to south. However, the advent of large-scale storage area networks and service-oriented architectures including resource pool management for virtualization and cloud computing has driven east-west traffic flow within the data center.

With deployment of clustered server structures based upon blade server architectures to support virtualization and cloud computing, applications which may initially be deployed in closely coupled server hosts can automatically, with high-availability applications, migrate to diverse hosts within a server farm. This can impact the number of switch hops between components of an integrated application.

In the diagram below, if two switches have the same “parent” in the hierarchy, then only three switches are involved in the connection. However, as the servers become more separated, the number of switch hops grow. Thus, a traditional tree switching architecture can lead to potential delay, risk of congestion, and packet loss. In a tree hierarchy, packets must first travel north-south before they can move east-west, which is highly inefficient.

To support the growing demand for east-west traffic flow within the modern data center, new network design approaches must be considered to optimize network design. Virtualization and cloud computing have had such a major impact on today’s data center traffic patterns that approximately 75% of traffic can be characterized as east-west. To optimize the performance of data center traffic, the traditional tree structure must be replaced with an any-to-any storage/server mesh design that reduces the number of hops north-south and delivers a single hop for flows east-west.
The logical solution to the data center switching problem is to achieve the concept of universal connectivity without introducing a deep switch hierarchy. This approach is called flattening the network and reduces the number of layers and weaves the remaining components into a common fabric that provides any port with dependable, high capacity connectivity to any other port within the data center facility.

To request more information on PTS Routing & Switching Design Services, click here.

Solution Showcase: Juniper QFabric Delivers Any-to-Any Connectivity within the Data Center
A true data center network fabric interconnects thousands of server, storage and other network ports in a flat, ultra-low latency infrastructure that provides any-to-any connectivity, ensuring that every device is a single hop away from every other device.

Juniper Networks QFabric technology delivers such a data center network fabric-a solution that flattens the entire network to a single tier where all access points are equal, eliminating the effects of network locality. The QFabric architecture also eliminates the need for protocols such as Spanning Tree, TRILL or SPB while reducing network latency to microseconds, improving the performance of every application in the data center.

By operating as a single device, the QFabric system-composed of the QFabric Interconnect, QFabric Director and QFabric Nodes-improves efficiency and simplifies operations by dramatically reducing the number of managed components from many to only one.

Juniper Networks Routing & Switching solutions are comprehensive, scalable, and secure solutions specifically designed to meet the needs of both enterprises and service providers. All of their routers and switches – run on one common operating system – Junos.

Juniper Router and Switching Solutions

PTS Education & White Papers
Check out the latest white papers and case studies added to our website by signing in or joining our new PTS Media Library! We’re sure there will be some interesting content for your reading pleasure. For up-to-the-minute educational information don’t forget to check out the PTS Blog – Data Center Design, consider joining the LinkedIn Computer Room Design Group, or visit our Data Center Design Channel on YouTube.

Cable Architecture Design for Next Generation Data Centers by PTS 

Each type of cabling architecture has notable advantages over the others relative to the data center’s function, density of equipment, up-time, and planned or unplanned growth. This white paper reviews the best approaches to data center cabling design and discusses when it is appropriate to use each. Download

Selecting the Optimal Data Center Cooling Solution by Peter Sacco, PTS

A better way to think about data center cooling is to forget the notion of adding ‘cold’ to a room… Further, selecting the optimal cooling solution involves a deep understanding and comparison between the performance characteristics, capital expense (CAPEX), and operational expense (OPEX) of potential configurations. Download

Datacenter Downtime: How Much Does It Really Cost? by Aberdeen Group

Aberdeen Group’s May 2010 research on Disaster Avoidance and Disaster Recovery: Making your Datacenter Disaster Resilient analyzed the frequency and associated costs of datacenter downtime for 125 organizations worldwide. The focus of this document is to quantify the cost of IT-induced business downtime. Download

Cooling Entire Data Centers Using Only Row Cooling by APC

Row cooling is emerging as a practical total cooling solution for new data centers due to its inherent high efficiency and predictable performance. Yet some IT equipment in data centers appears incompatible with row cooling because it is not arranged in neat rows due to the nature of the equipment or room layout constraints, suggesting the ongoing need for traditional perimeter cooling to support these loads. This paper explains how a cooling system comprised only of row coolers, with no room cooling system, can cool an entire data center, including IT devices that are not in neat rows. Download

Ten Errors to Avoid When Commissioning a Data Center by Schneider Electric (APC)

Data center commissioning can deliver an unbiased evaluation of whether a newly constructed data center will be an operational success or a failure. Proper execution of the commissioning process is a critical step in determining how the data center operates as an integrated system. The documentation produced as a result of commissioning is also the single, most enduring value added deliverable in a data center’s operational life. This paper outlines the ten most common errors that prevent successful execution of the commissioning process. Download

In the News
Some Data Center Operator Take Their Chances with Floods
Computerworld, July 16, 2013PTS President, Peter Sacco, was quoted in a recent Computerworld online article regarding how climate change is causing some IT leaders to consider relocating, or at least hardening, their facilities. Given the dire warnings about climate change, some business leaders and IT professionals are pondering this question: How should data center managers handle the crop of so-called 100- and even 500-year storms, coastal floods and other ecological disasters that climatologists predict are heading our way? Read the Article Data Center Renovation: 8 Warning Signs Your Data Center Needs an Upgrade
TechNews Magazine, July 2013Just like your home, data centers hit a point where slapping some paint  on the walls (or adding additional rack UPS or portable cooling systems (e.g. standing floor fans) to mask inherent design flaws, failures, or issues will no longer suffice. An investment in data center renovation will keep your data center running much more smoothly and will help minimize the chance of a disaster. Consider the following warning signs, and what you can do to fix each issue, to keep your data center shiny and new. Read the NJTC TechNews ArticleErnst & Young Names Pete Sacco Entrepreneur Of The Year® 2013 Finalist in New Jersey Press ReleaseMay 3, 2013
Franklin Lakes, NJ, May 3, 2013 – Pete Sacco, President & CEO of PTS Data Center Solutions (“PTS”), was named a finalist for the Ernst & Young Entrepreneur Of The Year 2013 Award in the New Jersey Region. The awards program recognizes high-growth entrepreneurs who demonstrate excellence and extraordinary success in such areas as innovation, financial performance and personal commitment to their businesses and communities. These finalists were selected by a panel of independent judges. Award winners will be announced at a special gala event on Thursday, June 13, 2013 at the Hyatt New Brunswick.

Read the Full PTS Press Release

About PTS Data Center Solutions

Founded in 1998, PTS Data Center Solutions is a premier data center design firm and turnkey solutions provider, offering a broad range of project experience, specializing in designing data centers, computer rooms and technical spaces that integrate “best-of-breed”, critical infrastructure technologies and result in continuously available, scalable, redundant, fault-tolerant, manageable, and maintainable mission critical environments.