A Network Fabric for New Applications in the Data Center

The Network Challenge with Deploying Applications
Thanks to virtualization organizations can bring up new applications and services quickly. Unfortunately, many data center networks don’t let you fully capitalize on the business agility that virtualization and modern application architectures provide. Traditional network architectures are too slow and too cumbersome to configure. For true agility, enterprises need a high-performance, low-latency fabric network that can be managed like a single, logical switch.

Today’s mid-sized and large-scale data centers are built with high-performance blade and rack servers, which typically run multiple VMs, which in turn run increasingly modular, web- and cloud-based applications. These modern applications are driving increased traffic levels and different traffic patterns, which place specific demands on the data center network.Critical business applications such as enterprise resource planning (ERP) and customer relationship management (CRM) are divided into multiple modules. As a result, relatively simple operations such as entering an order can trigger 100 or more individual transactions, each imposing its own latency. Some applications, including eCommerce applications, even behave dynamically, spinning up new instances or migrating workloads in response to traffic loads.

Their distributed nature means modern applications are spread across racks of servers, each served by multiple switches. The applications generate a tremendous amount of server-to-server, or east-west, traffic as the various modules communicate with one another. Multi-tier network architectures aren’t well matched to modern applications. They force this east-west traffic to first travel north and south, up and down the network tree, before arriving at its ultimate destination, adding significant latency that can cause application performance to degrade under load.

The Critical Role of the Business Edge Network

The Service Provider’s Challenge
Businesses are increasingly using service providers to host critical applications and data in order to better control the security and availability of the data and to mitigate the expenses associated with hosting and serving data locally to the entire business user base. They access these applications and data over the service provider’s business edge network. This creates a challenge for the service provider, because their customers measure the success of the network by its ability to handle critical data and provide a superior user experience. With more and more applications and data centralized, the importance of the network, and its role in the success of the business, becomes ever more critical.

Market forces have created an environment of unpredictability for the service provider. Mobile devices, new applications, and the flood of new content place a direct strain on the business edge network, forcing the business and the service provider to address the network in creative ways. Maintaining and evolving the end-user experience is critical to the success of the network and key to enabling the service provider to meet its business goals.

The business user base, places considerable expectations on the network. Users expect their business e-mail and voice over IP calls to work flawlessly, while also expecting mobile video and applications to perform seamlessly. The impact of network performance is real and must be factored into the design of future business networks. This challenges the business edge provider to optimize the design and find ways to reduce the cost in the face of the increasingly complex and growing demand for a high-quality end-user experience.

Leveraging Data Center Interconnect For Business Continuity

It’s Time to Reengineer the Wide Area Network
Similar to the data center, organizations are growing increasingly reliant on the wide area network that connects data centers to help them run their businesses.  While the WAN is key to data backup plans for maintaining business continuity, many WAN links aren’t up to the task. Standards-based data center interconnect technology could help you keep operating in the face of a disaster, but some reengineering of the WAN might be in order if the plan is going to work.

Business continuity planning is not always a top priority, since many view it as insurance and not a business driver.  However, you might be able to combine efforts to help you achieve your goals. For instance, transitioning to a private cloud may require WAN links to be reengineered to provide bandwidth on demand so that virtualized workloads can move between data centers based on user demand and resource utilization. This has the added benefit of supporting the varying bandwidth requirements for backing up data between data centers.

Business Continuity and DCI Strategies
When you use DCI to replicate data between geographically distributed locations you will want to configure your LAN connections so that application and storage traffic can flow between data centers as needed to so that you can maximize application availability and provide data redundancy in the case of an outage. Here are some things to think about.

Making the Transition to Converged Storage and Data Networks

Storage and data network convergence holds the promise to transform the data center and make it a more cost effective operation for the Enterprise. There is the potential for considerable savings as a result of reducing the number of network interface cards per server, reducing cabling, and lowering the power and cooling draw, as well as having one less physical network to manage. The change is made possible by the capability to transport Fiber Channel frames over 10 GB Ethernet using Fiber Channel over Ethernet technology. Making the transition isn’t an easy task though. Let’s take a look at some of the considerations and how the transition can be made more easily.

The value of converging networks using FCoE is compelling and many organizations are considering making the move to FCoE but the question is how to do it without disrupting operations. For organizations that are building new data centers and consolidating older ones the answer is easy. They can just build an FCoE capable network in the new data center and migrate their applications and storage over to the new infrastructure. The more difficult situation is what to do if you need to convert an existing production network and cut over to FCoE live? This is where it gets interesting.

A Practical Look at SDN Software and Hardware Considerations

The Opportunity for SDN
Software Defined Networking represents the biggest change to the network in many years. What makes SDN interesting is the transformation that it can enable. Businesses are looking for more control over their applications on the network. SDN promises to deliver agility and simplification in the network to support applications. With SDN, the network becomes more efficient and agile, and an enabler for delivering on business goals for application performance. As a buyer, it’s understandably difficult to separate the hype from the reality. I’d like to suggest a few points to consider as you map out your SDN strategy.

Implications for the Network
Network operators still need to design, provision, manage, operate and troubleshoot their network. While SDN offers greater simplicity, operators have to continue current network management functions and, at the same time, become educated on  SDN developments. With SDN, new protocols and technologies will emerge. The investments you make in your network infrastructure today need to be flexible enough to see you through the next several years. There will be a hybrid model in the network. This means a mix of overlay technology and physical networks – and the demarcation points will depend on use cases for those overlays and the ability for the physical network to support these overlays. Network operators will have to understand the relationship between the two and be able to design networks appropriately with the right on-ramps and off-ramps. As the management of network shifts away from CLI and more towards orchestration platforms, the network interfaces and integration points (APIs) need to be clearly defined and deliver automation and agility.

Introducing Juniper's Enterprise Data Center Solution

The data center infrastructure is evolving. Application architectures have become more distributed; compute has gone from dedicated hosts to hypervisors, and storage has become converged and shared.  The challenge facing organizations is that they need to evolve the network to better serve the business. The network must become simpler and flatter, more automated, more open and integrated with the rest of the environment. A simple and flat network delivers the performance, and management and orchestration integration help the businesses to stay agile. Let’s take a look at how we make this happen.

Challenges in the Data Center Today
The two major challenges in today’s virtualized data center networks are connectivity and operations. Organizations are looking to achieve greater business agility to respond and adapt quickly to business needs. For enterprises that view IT as a strategic business asset, this is characterized by:

Automate Network Configuration with Puppet for Junos OS

With the exponential growth in the numbers of servers in Enterprise Data Center and the corresponding complexity and confusion that can follow the need for management automation of server resources is well understood by the IT staff. However for every server there are one or more network connections that need to be configured and managed and until recently this was still a manual process.  Juniper Networks has addressed this challenge with the release of Puppet for Junos OS.

Why Resource Automation is Needed
Driven to reduce costs while providing high levels of computing power, enterprises are adopting cloud-based computing models based on large numbers of low-cost processors and virtual machines. This is fueling an exponential increase in the volume and complexity of server and network management. Meanwhile, understaffed IT organizations are tasked with supporting the business, as the organizations they serve are under pressure to move with agility in a competitive business environment.

Build your Network with the Vertical Campus Implementation Guide

Designing a large campus network can be a daunting task. It is certainly a complex undertaking. It can be especially difficult if you have to configure equipment that you haven’t worked with before. You might find yourself wishing that someone would write a how to guide. At Juniper Network we hear you and that’s why we’ve published another one of our Implementation guides. This one is for the Vertical Campus. It will show you the way to set up Juniper LAN, WLAN, and security devices and help you get up and running in a shorter time, with a tested and proven design.

The Vertical Campus Implementation Guide
The Juniper Networks® Vertical Campus Implementation Guide provides a simple, tested, step-by-step process for rapidly deploying a large campus solution. The design incorporates the most commonly used enterprise network technologies to provide a simple and scalable network architecture that includes LAN, WLAN, and security components. The guide presents a specific configuration of Juniper Networks hardware and software platforms that have been tested and provide a reliable foundation on which to base a customized network for your business.

Inter Data Center Workload Mobility with VMware

Server virtualization has increasing become implemented in the data center because it enables higher utilization of physical servers increasing their value to the organization. By abstracting the server operating system from the server hardware, virtualization allows physical resources to be shared among many virtual machines. The capabilities of cloning, suspending, and migrating live VMs among a cluster of physical hosts enhances resiliency and performance of applications. Advancements in networking technology allow servers within a single cluster to be located across the entire data center, or in another data center some distance away, further enhancing application availability, but making this technology work isn’t simple.

The Case for Live Migration
There are may use cases put forward for live workload migration and they are covered in the document, and were also covered in my previous blog “Making the Case for Long Distance Virtual Machine Mobility.” These use cases include optimizing server resource utilization, optimizing resource consumption at various locations, hybrid cloud where overflow workloads move to another data center, and disaster avoidance were workloads are moved to saftey, as well as a follow the sun model where workloads move according to the time zone of users. Some scenarios for data center migration for live workloads can be accomplished without implementing complicated first-hop-responder and route-optimization techniques. Other scenarios are possible however, they require implementing protocols that share state and fate, thereby increasing the complexity and reducing the reliability of the connected data centers. As a result of these considerations the network is a critical factor.

Network Functions Virtualization is Changing How Services are Delivered

As Service Providers are faced with increased competition from Over the Top Providers they are seeking new markets to enter. However, as they look to create and launch new services they must grapple with the growing number and complexity of hardware devices in their networks. This creates challenges due to the time it takes to certify equipment and with staffing and training of skilled operators for many devices. It also creates cost pressure with the need for more space and power at a time when these resources are becoming ever more expensive. Making upfront capital outlays for equipment in anticipation of revenue that ramps up over time can stress budgets. As a result Service Providers are looking to change how network services are deployed and some are finding Network Functions Virtualization (NFV) as the answer to their problems.

Defining Network Functions Virtualization Services
Network Functions Virtualization is transforming how network operators architect networks by enabling the consolidation of network services onto industry standard servers, which can be located in Data Centers, on Network Nodes or at the user premises. NFV involves delivering network functions as software that can run as virtualized instances and that can be deployed at locations in the network as required, without the need to install equipment for each new service. Network Functions Virtualization is applicable to any network function in both mobile and fixed networks. Network Functions Virtualization is complementary to Software Defined Networking (SDN) but not dependant on it. Virtual appliances might be configured using SDN capabilities and they might be connected via overlay network tunnels in clusters based on an application or based on the needs of an organization.

Achieving Single Touch Provisioning of Services Across The Network

Service provider networks are continually growing and evolving in order to provide a rich end user experience. Ethernet-based services are increasing rapidly to deliver high bandwidth and anytime access to video, voice, and broadband applications. In order to meet the needs of their customers, service providers have built complex networks consisting of thousand of high capacity Ethernet ports on devices in many locations. Typically these devices need to be provisioned, configured, and managed separately, which increases operating costs and time to deploy new services. The challenge is how to achieve single touch provisioning of services across the network.

Most factors contributing to the complexity of service provider networks revolve around the need to provision and manage separate physical platforms. The majority of the service provider’s OpEx cost results from the necessity to have the technical support infrastructure required to maintain and deploy services in a distributed network. Case studies show that by eliminating the need to manage individual devices separately can reduce the service provider’s OpEx by 70%.

Juniper Networks understands the challenges faced by service providers in today’s environment and has created a system called the Junos® Node Unifier (JNU) that enables centralized management, provisioning, and single touch deployment of services across thousands of Ethernet ports. The JNU solution is Juniper’s way to simplify today’s networks. JNU consists of satellite devices connected to a hub, where the satellites can be configured, provisioned, and managed from the hub device, giving a single unified node experience to network operators.

OpenStack Summit San Diego 2012

The OpenStack Summit in San Diego was the place to be last week. I attended and wanted to share my observations. There was a lot of participation and energy. It was sold out with over 1300 people attending and about 35 vendors displaying their products.  Previous summits were developer forums. This time the format was expanded and there were hundreds of sessions in many categories including case studies, industry analysis as well as the usual develop sessions. See, http://openstacksummitfall2012.sched.org/ for a list of the session.

What is Openstack and What Does it Do?
OpenStack (www.openstack.org) is a foundation that manages an open source cloud computing platform. It was founded in 2010 out of a project that was started by NASA and RackSpace to build their cloud infrastructure. Their mission is “To produce the ubiquitous open source cloud computing platform that will meet the needs of public and private cloud providers regardless of size, by being simple to implement and massively scalable.” OpenStack focuses on the core infrastructure for compute, storage, images and networking. There is a large ecosystem of vendors providing tools do the things that OpenStack does not do.

OpenStacks consists of modules to configure cloud computing resources. The components are the Nova Compute Service, Swift Storage Service, Glance Image Service and Quantum Network Service. They automate the functions that are required to set up these services. OpenStack lets organizations quickly provision and reprovision compute resources. Even with virtualization it can take days to fully set up a virtual server, networking and storage. Organizations want this to happen in minutes whether they are offering a commercial service or an internal IT service.

Have We Hit The Network Breaking Point?

We have all noticed that the business environment is changing at a rapid pace. IT departments, previously seen as cost centers, were tasked with reducing their infrastructure and even justifying their existence, but not anymore. With the rise of just-in-time manufacturing, and long tail customization, all driven by ecommerce that’s increasingly accessed from mobile devices, the old rules are being turned upside down.  The IT organization is now tasked with enabling the business to drive innovation.

The Opportunity that Comes with Change
While change comes with opportunity, we know change isn’t always easy. The old ways of building infrastructure and networks are no longer working. Organizations need to adapt to a world where customers use social media and create huge amounts of data that contains business intelligence.  All of us have questions about what is happening in the IT space, what the challenges are and what do to about it. That is why Juniper got together with Forrester and did some research to help sort this all out.

It started in July 2012, when Juniper Networks commissioned Forrester Consulting to evaluate what enterprises need from a network in order to scale for business and meet future needs. Forrester Consulting surveyed 150 IT business decision-makers from enterprises across the United States, and yesterday the results were shared in a webinar. It was a conversation between Forrester Analyst Andre Kindness, Juniper’s CIO Bask Iyer and Mark Settle from BMC Software. I listened in and there was a lot to learn. 

Monetizing the Business Edge with Hosted Private Cloud

The last few years have seen major changes in enterprise networking, with the growth of MPLS virtual private networks (VPNs), and the rise in adoption of cloud services. A growing trend, which blends the two, is Hosted Cloud Services, delivered over MPLS VPNs from the service provider’s data center to the customer’s branch office locations. These services are facilitated and enhanced by Juniper’s combined routing and switching solutions as I discussed in my blog about the Junos Node Unifier.

Network Service Providers’ Opportunity with Cloud Services
NSPs are looking at new areas to boost revenue growth, and cloud computing is one of the key areas they are looking at. Based on research from STL Partner’s Telco 2.0 report, the Hosted Private Cloud market is expected to grow 34% annually to $6.5 billion by 2014. Since they already own the network, as well as the infrastructure around it including the billing and management systems, NSPs are in a good position to monetize the cloud opportunity. While the cloud services market is highly competitive, there is an opportunity for NSPs to differentiate their service offerings by leveraging their enterprise-grade VPN infrastructure to provide Hosted Private Cloud services with enhanced end-to-end security and service-level guarantees.

Public cloud services lack the SLA and QoS guarantee levels that enterprises have grown accustomed to with their VPN networks. Recent power outages associated with major public cloud service providers have impacted many popular sites and highlighted issues associated with reliance on public cloud services. As a result hosted private cloud services are emerging as a cost-effective and robust solution that offers quality and reliability for enterprise applications. Network service providers (NSPs) are embracing cloud services to grow new revenue streams and increase customer retention.

Innovating at the Edge in the Age of the Cloud

With the rise of Software as a Service and Social Media network service providers are witnessing a game changing shift in how consumer and business services and applications are delivered. Some service providers see the opportunity to break beyond their connection oriented business model and embrace these new cloud-based services. In order to do so they are looking for ways to adapt their networks to accommodate these new services.

Service Delivery on the Edge
In order to take advantage of the new business models brought about by the service transformation network service providers need to consider a number of factors, including adopting progressive business and monetization strategies and considering subscribers’ preferences in the service definition process. For the network service provider the most critical change is to leverage the underlying network architecture to support the new service offerings. As a result the traditional architecture of service provider edge networks is undergoing an aggressive period of evolution and shifting focus from simply a point of network connectivity to becoming a vital services creation and innovation point.

Subscriber Defined Services
Matching services and applications to customer expectations has always been a formidable challenge for telecom operators. However, it is now even more difficult given that the expectations of a telecom subscriber have changed drastically over the past few years. In the past consumer and business subscribers were tethered to the network services provider as their sole source of services, but today these subscribers have connections to OTT providers and access a variety of personal and business applications. An Important change is that these subscribers are not just service consumers, but they are also shaping service innovations by leveraging more intelligent and programmable platforms and devices.

Solving the Network Services Provisioning Challenge

Centralization of applications at large scale in a consolidated data center, and the increasing size of software-as-a-service application deployments, create the need to deploy thousands of 1GbE/10GbE ports to connect to servers and storage devices in the data center. Service providers need a way to terminate application servers without having to build layers of separately managed networking devices. They also need to be able to configure services from a central location and automate the provisioning of their network devices. This is driving the need for a device-based solution that can control thousands of network ports from a single point and that can interface with service orchestration systems.

The Services Provisioning Challenge
As service providers seek to deploy new cloud and network services at high scale, managing and maintaining individual network devices adds additional layers of operational complexity. As layers of network devices are added to the environment, service providers have to work with multiple management systems to provision, troubleshoot, and operate the devices. These additional layers may translate into additional points of failure or dependency reducing service performance. To meet this challenge Juniper Networks has released the Junos Node Unifier, a Junos OS platform clustering program that reduces complexity and increases deployment flexibility by centralizing management and automating configuration of switch ports attached to MX Series 3D Universal Edge Routers acting as hubs. You can visit the landing page for the product launch here, link.

Simplifying the Network
The Junos Node Unifier enables scaling up of applications in the data center by supporting a low cost method to connect network devices to a central hub. It reduces equipment and cabling costs and increases deployment flexibility by centralizing management and automating device configuration, while overcoming chassis limitations to enable the connection of thousands of switch ports to be attached to the MX Series platform. The Junos Node Unifier solution leverages the MX Series modular chassis-based systems as well as access platforms including Juniper Networks QFX3500 QFabric Node, Juniper Networks EX4200 Ethernet Switch and EX3300 Ethernet Switch, to be used as hub and satellites respectively. Junos Node Unifier leverages the full feature set of these devices to support multiple connection types at optimal rates, with increased interface density as well as support for L2 switching and L3/MPLS routing on the access satellites.

Getting on the Path to SDN with Juniper Networks

A number of IT trends—including the consumerization of IT, cloud computing, and social media— present significant opportunities for businesses to improve productivity. Before adopting these technologies, however, organizations try to fully understand the impact they will have on the underlying infrastructure and, more specifically, the network environment, since it is a critical enabler for all of these services. As a result organizations are looking to innovation in the network to meet their business needs.

Trends Impacting the Network
As organizations continue to innovate and expand their virtualized environments beyond the simple benefits of consolidation to a more agile infrastructure, they have begun to build out private clouds. These agile IT environments enable business managers to rapidly turn up new services to meet unexpected demand or requirements. However, this abstraction layer can create blind spots in the infrastructure and make meeting compliance requirements difficult.

Social media applications present another challenge as the explosion in the number of network-connected devices opens up avenues to new applications and collaboration tools. Well known applications such as Facebook, YouTube, and Twitter often blur the lines between business and personal usage and video create demands on the network. The question is how to ensure performance, while protecting data ensuring privacy.

Juniper’s Internet Edge Implementation Guide is Here

Juniper has created an implementation guide that will help network designers create a simplified Internet edge solution using Juniper Networks MX Series 3D Universal Edge Routers, SRX Series Secure Services Gateways, and EX Series Ethernet Switches. This guide details specific design considerations, best practices, and Juniper tools that can be used to build the optimal solution. It concludes with a real-world deployment example that illustrates the solution and recommended configurations in detail.

The Role of the Internet Edge
The Internet edge acts as the enterprise’s gateway to the Internet. It provides connectivity to the Internet for data center, campus, and branch offices, and it connects remote workers, customers, and partners to enterprise resources. It can also be used to provide backup connectivity to the WAN for branch offices, in case the primary connection to the enterprise WAN fails.

Today’ s Internet edge must enable access to a variety of applications such as cloud computing solutions, mission critical applications, and bandwidth hungry applications such as video. The Internet edge must also scale seamlessly to support growing application performance and bandwidth needs, while supporting a rich set of routing and security features. This guide will help you reach this goal.

The Myth of the Broadcast Domain Apocalypse

There is a train of thought that is popular with some network vendors and long time network engineers that there is a compelling need to “orchestrate” the physical network and the visualized-networking world.  This is expressed as a desire to ensure that VLANs are synchronized between the physical network switches and the virtual switches on the visualization hosts.

The Orchestrated VLAN Model
VMware has a concept of a “backing VLAN”. This simply put means that traffic belonging to a portgroup uses a configured transport VLAN when it traverses the physical network.  For example, if a group of VMs belong to a portgroup “backed” by a VLAN, that VLAN must be “allowed” on the trunk ports connecting all of the physical ESX hosts that are hosting any of those VMs.

In addition to this, it is argued that the trunk ports to ESX hosts that do not host a VM belonging to that portgroup should not “allow” the presence of the VLAN that backs that portgroup.  It is suggested that the reason for this requirement is that the unbridled propagation of VLANs will cause the ESX host to process broadcast packets it does not need to, with potentially dire consequences.

Why Network Latency Matters for Virtualized Applications

My colleague Russell Skingsley has an interesting take on the effects of latency on virtualized application performance. The purpose of virtualization is to optimize resource utilization. As Russell pointed out it isn’t just an academic conversation. For the cloud hosting provider it’s about revenue maximization. Network latency has a direct impact on virtualized application performance and therefore revenue for the service provider. Your choice of network infrastructure will impact your business, but it can be difficult to see how investing in high performance networking solutions will improve your business. I will try to connect the dots.

Don't Sell it Just Once
It’s a service provider axiom that you don’t want to waste your valuable assets by selling them to a single customer, even at a premium. Take dark fiber for instance, every SP has come to learn that selling a fiber to a single customer will never bring scalable revenues. Making the move to adding DWDM to the service mix allows you to scale up the number of customers and is an improvement over selling dark fiber, but is limited by the number of lamdas that can fit in the spectrum on a single fiber. The scaling is linear. A similar situation exists for shared cloud computing resources.