A CIO s Guide to Hybrid Cloud Performance

WHITE PAPER A CIO s Guide to Hybrid Cloud Performance How Riverbed Accelerates Public and Private Clouds: A Framework for Hybrid Cloud Adoption

Introduction: Hybrid Cloud Strategies - Driving the IT Roadmap The advent of cloud computing offers new choices in architecting IT infrastructure for the best possible blend of performance, availability, cost, and control. As a result, there will be both consolidation to fewer data centers and migration of selected applications and data to more cost efficient public cloud services. A critical requirement for this transition is the ability to integrate public and private services with optimized connectivity to build a flexible hybrid cloud environment. Riverbed enables this business agility with a comprehensive range of high performance solutions for the hybrid cloud including WAN optimization, network performance management, application delivery controllers, and cloud storage gateways. Applications and data can be accelerated, managed, and secured, delivered from the public cloud, to the data center, to the branch, and to users anywhere. Most organizations have three basic goals: 1) increase profits or productivity, 2) reduce costs, and 3) manage risk. IT continually works to support these goals, but there are often conflicts and trade-offs: highly secure and protected systems are expensive and can hamper user productivity; new, revenue-generating initiatives often bring new costs and management challenges, etc. A mixed architecture of public and private clouds can help IT address all three goals simultaneously. Increase profits or productivity With the pace of business driving ever forward, IT is challenged to support an agile infrastructure. The capital costs and procedural bureaucracy of new application development in IT organizations inhibit innovation, which is essential to discovering and supporting new revenue streams, and in turn drive profit expansion. Meanwhile, those new applications and services require seamless delivery in order to maximize user productivity. Private and public cloud models minimize or eliminate the capital costs to develop new applications and services. Key enabling technologies, such as WAN optimization, network performance management, application delivery controllers, and web content optimization ensure applications perform without inhibiting user productivity. Reduce costs IT procurement has driven over-capacity buying for three primary reasons: 1) to support a distributed, underutilized infrastructure that meets performance requirements; 2) to support peak demand for resources, and; 3) to adequately protect systems with many layers of redundancy. Private and public cloud models support consolidation of resources, driving out underutilization inefficiencies of a distributed environment; efficient, on-demand provisioning of resources, reducing the need for redundant, pre-provisioned capacity; and may allow for inter-cloud brokering models to further reduce per unit costs in the future. Manage risk Unprotected systems and mismanaged, distributed architectures increase the risk of a negative impact on the business from data loss and system failures. Private and public cloud models support securely consolidating, replicating and automating systems management for reduced risk of unavailability, while also improving the cost and efficiency of broader system protection. Key Enabling Technologies for Hybrid Cloud Performance Ø WAN Optimization controllers provide symmetrical solutions to improve the performance, efficiency, deployment and security of a wide range of applications. Ø Network Performance Management consists of measuring and monitoring networks to ensure that they carry traffic with the speed, reliability, and capacity that is appropriate for deployed applications. Ø Application Delivery Controllers, often described as next generation load balancers, provide asymmetrical solutions to improve the performance, efficiency, deployment and security of a wide range of applications Ø Cloud Storage Gateways are network appliances or servers which reside at the customer premises and translates cloud storage APIs such as SOAP or REST to block-based storage protocols such as iscsi or Fibre Channel or file-based interfaces such as NFS or CIFS Ø Web Content Optimization is software that reduces the time for a browser to load information from a website or intranet. This paper outlines a framework and specific solutions for adopting a hybrid cloud architecture to support increased profits and productivity, reduced costs, and improved risk management. Specifically, this paper seeks to answer a number of questions related to three phases in the journey to high-performance hybrid cloud computing, including: Ø Strategic Planning: What are the different cloud models available? What challenges must be overcome to adopt the various cloud models? What do organizations need to do to prepare to adopt cloud computing? Ø Flawless Execution: What are the challenges to migrating and deploying in the cloud? How can administrators overcome distance challenges and application-specific challenges? How can organizations take full advantage of cloud scalability in hybrid environments? Ø Seamless Delivery: What are the challenges to delivering and managing cloud services? What systems facilitate meeting SLAs? How can organizations accommodate change in their cloud deployments? How do organizations need to adapt their data protection and business continuity strategies for the cloud? 2011 Riverbed Technology. All rights reserved. 1

Phase One: Strategic Planning The key to successfully adopting and leveraging cloud computing is understanding the existing IT environment and making informed decisions about which services to move to a public cloud provider, which pieces of infrastructure to centrally manage in a private cloud data center, and which technologies to keep distributed with users or in branch offices. The biggest challenge for many organizations is determining how they can best adopt a service delivery model that gives them the best possible blend of performance, cost efficiency, and reliability. To do that, IT decision makers need to consider the challenges associated with hybrid cloud models, as well as their own environment. Understanding the Hybrid Cloud Cloud computing itself takes many forms. Public, private, and hybrid cloud computing models have some fundamental characteristics in common such as a consolidated and virtualized platform, elastic on-demand scalability and provisioning, metered cost accounting, and high availability and resilience. They do however also have some drastic differences regarding maintenance and accountability for the systems, elasticity of shared resources versus security concerns, and capital versus operating costs: Ingredients for Strategic Hybrid Cloud Planning: Ø Network Performance Monitoring (NPM) Ø Application Dependency Mapping 1. Public cloud offerings which connect the organization to cloud service providers built with shared infrastructure components, 2. Private cloud environments where organizations build their own consolidated data centers, and 3. Virtual private cloud environments where service providers offer dedicated infrastructure. Hybrid clouds include a combination of cloud models, which must replace or integrate with conventional, pre-existing IT environments. Since few organizations expect to move entirely to a single cloud model, or to avoid adopting cloud computing completely, hybrid clouds appear to be the model that most organization will adopt, if they are not already there. GigaOm recently noted, Any survey of IT executives will show hybrid clouds as a top priority, and ours is no different, with 69 percent of respondents indicating an expected shift to a hybrid model over time. Hybrid clouds are popular because they give organizations the freedom to use public and private cloud resources in conjunction with one another without having to commit entirely to either model. 1 Operating in hybrid cloud architectures require that IT surmount a number of challenges. These challenges are familiar they are based on existing IT challenges but they have unique implications related to the consolidation of IT resources: Hybrid clouds are popular because they give organizations the freedom to use public and private cloud resources in conjunction with one another without having to commit entirely to either model. 1 GigaOm, A field guide to cloud computing: current trends, future opportunities, June 2011 1 - GigaOm, A field guide to cloud computing: current trends, future opportunities, June 2011 2011 Riverbed Technology. All rights reserved. 2

Ø Visibility Consolidating, centralizing, and migrating applications to third-party cloud providers all require knowledge of application dependencies before making these transformative changes to an environment. Managing service performance across public and private clouds requires application-aware visibility for rapid troubleshooting. Ø Distance Moving applications into public clouds or centralized private data centers both introduce distance and, therefore, latency, which hinders performance. Ø Data Data is forecasted to continue to grow at a rapid pace, with IDC estimating storage to grow 45x between 2010 and 2020 2, regardless of whether in the public or private cloud. In addition to adding congestion to both Internet and private WAN-links, protecting increasing amounts of data in a cost efficient manner is also a challenge. Ø Access Most organizations backhaul Internet traffic from mobile users and branch offices. As public cloud services become increasingly adopted, IT will bear the burden of managing service delivery to remote users, in addition to traditional WAN-based corporate applications. Ø Security and compliance Some data sovereignty regulations may require data to reside in specific regions, which limit which public cloud services are available as options. Organizations need to understand and work with cloud service providers to ensure that security requirements, such as data center audits and authentication services, can be met. Understanding Existing IT Deployments Major IT projects such as IT consolidation and cloud deployments have one thing in common: the need to discover what s running on the network, and who is using it. The discovery phase of IT projects has historically been time-consuming, expensive and errorprone, with several months spent scanning the network, finding dependencies and updating asset lists. Regardless of whether the plan is consolidating a few servers to public cloud or moving entire data centers to a private cloud model, Riverbed can help accelerate the project, achieve planned savings sooner and reduce the risk of outages and failed cutovers. Within a few hours, Riverbed can discover all systems, identify server and client dependencies and locations, and show the information in a graphical, interactive format. Phase Two: Flawless Execution After identifying which applications are candidates to centralize into a consolidated private data center and which are candidates to move to a public cloud service, the next set of challenges are successfully migrate data and applications, as well as ensuring access to those resources: Ø Migration - Migrating data across the WAN or public Internet is Ø Application Delivery Controller (ADC) negatively impacted by latency. In addition to the time and Ø Web Content Optimization (WCO) complications of potentially restarting migrations or synchronizing data, the cost of additional bandwidth to private cloud facilities and data transmission fees into and out of public clouds can become prohibitively expensive, depending on the distance and volume of data to migrate. Furthermore, data migration competes with other applications for network resources, threatening to disrupt application performance broadly if left unchecked; Ø Performance - Users experiencing slow responsiveness to data or applications will suffer productivity losses when using applications or data that they already depend on and then will be less likely to adopt a new services; Ø Scalability Whether building a private cloud infrastructure that needs to span resources in multiple data centers or building applications to take-advantage of public cloud capacity, developing an architecture that remains agnostic to location and minimizes incremental investment is essential to building a scalable hybrid cloud. Optimizing migration of hybrid cloud workloads and data Ingredients for Flawless Hybrid Cloud Execution: Ø WAN Optimization (WOC) Migration to public and private clouds is challenged by network latency, cost, and contention all of which can be addressed by a WAN optimization solution. Riverbed s Steelhead family of WAN optimization products use data, transport, and application streamlining to eliminate redundant WAN traffic by de-duplicating data, improve the performance of all TCP-based applications, and decrease the effects of latency by reducing the number of application-level round trips across the WAN. Whether deployed as an appliance in branch offices and private data centers or as a virtual machine in private or public clouds, Steelheads optimize the migration of workloads and data to the hybrid cloud in two stages: 2 IDC s The Digital Universe Decade Are You Ready, May 2010 2011 Riverbed Technology. All rights reserved. 3

Ø Primary migration: Whether transferring applications, data, or even full server images from branches to centralized data centers, or from private data centers to public clouds, the Steelhead solution can accelerate the initial transfer; Ø Extended transition: Many organizations get started with a period of testing time. If the testing is successful, the legacy deployment is decommissioned and the cloud deployment becomes primary. During the testing phase and after the initial migration, the Steelheads will continually optimize communications between users on your network and applications running on instances in the cloud, providing LAN-like performance. Optimizing performance of hybrid cloud applications Applications, include file sharing, email, and more, are critical to business productivity and increasing revenues. As IT organizations look to deploy or migrate applications into the hybrid cloud, Riverbed optimizes the performance and quality of these applications for workers anywhere. In addition to WAN optimization, Riverbed web content optimization can accelerate page load times and reduce data transfers for web-based applications, such as SharePoint. Riverbed global load balancing technology may be used to route users to the most proximal and available instance of an application, further improving page load times and enhancing user experience. With improved application and website performance, organizations can drive increased user activity and revenue. Examples of applications that benefit from Riverbed s powerful application-aware optimization technologies include: Ø Microsoft SharePoint Server. Steelhead s SharePoint optimization incorporates all HTTP optimizations plus a deep understanding of the specifics of SharePoint traffic. These extend to HTTPS, too, which is crucial when the servers are accessed over a public network, as is the case with SharePoint running in the cloud. Stingray Aptimizer extends Riverbed optimization to the SharePoint pages themselves, which are dynamically tuned for faster page load times. Ø Microsoft Exchange Server. Steelhead s MAPI optimization works for both encrypted and unencrypted sessions between Outlook and Exchange. Users benefit from the full suite of optimizations, including data reduction and application protocol acceleration. Ø Web-based applications, including ERP/CRM. Any application that uses a web-based front-end to communicate with a database will benefit from the improved performance and responsiveness provided by Steelhead s HTTP optimizations. Beyond the standard HTTP optimizations, Steelhead provides additional acceleration of SAP and Oracle deployments. This is especially important for organizations that choose to deploy hybrid cloud models, because such deployments often involve continual transfers of large amounts of data between cloud and on-premise resources. Further enhancing the availability, performance and security of web applications, Stingray Traffic Manager s auto-scaling, caching, offloading, and request inspection capabilities enable applications to meet peaks in demand and respond more quickly, while isolating components of the application. Ø Software-as-a-Service (SaaS). SaaS offerings can be improved via partnerships that extend optimization end-to-end from the cloud service provider across the Internet all the way to the end users on the WAN. Current Riverbed WAN optimization offerings have customers using services such as Microsoft Exchange Online or SharePoint Online with no performance penalty, and future joint solutions with Akamai will incorporate acceleration for almost any SaaS application or provider. Optimizing hybrid cloud scalability Continuously adding capacity across a widely distributed infrastructure is no longer seen as a cost effective approach, but combining available private cloud capacity and accessing the virtually infinite capacity on demand in public clouds is challenged by integrating these varied resources and performance across distances. Riverbed hybrid cloud solutions allow organizations to access resources in multiple locations for enhanced scalability: Technology Spotlight: Riverbed Optimization System (RiOS) Ø Data streamlining. Eliminates redundant WAN traffic by de-duplicating data in all TCP flows. Data streamlining analyzes all traffic into indexed segments and stores those, along with references, in the local cache. When RiOS has seen a pattern once, during subsequent accesses it sends only the references. Each appliance, whether physical or virtual or in the cloud, has a single universal data store that s shared across all applications. Ø Transport streamlining. Improves the performance of all TCP-based applications by adapting transmission characteristics such as window scale, loss handling, and congestion notification. Transport streamlining reduces the number of TCP packets required to complete a transaction, including those protected by SSL. Ø Application streamlining. Decreases the effects of latency by reducing the number of application-level round trips across the WAN. Application-specific optimizations remove much of the chattiness and overhead associated with common applications that normally behave fine on LANs, giving these applications a local-like feel even though they re running at a distance. 2011 Riverbed Technology. All rights reserved. 4

Ø Balancing: Stingray Traffic Manager contains is an integrated management solution that gives a coordinated view of application delivery fabric, that may span multiple, geographically distributed, datacenters and cloud-based environments. Ø Cloud Balancing: To minimize risk of failure of a single cloud, and to provide DR capability, Stingray Traffic Manager can balance requests across multiple clouds based upon proximity and availability Ø Cloud bursting: For workloads that largely remain in the data center, but leverage public cloud resources during peaks in demand, Stingray Traffic Manager s auto-scaling capability can control on-demand provisioning and scaling of application instances that have been preloaded into the cloud data center. Phase Three: Seamless Delivery With the rise of cloud computing, the role of IT increasingly shifts to that of an internal service provider, with business unit customers demanding high-quality service, sometimes with service level guarantees. Riverbed hybrid cloud solutions offer significant improvements to IT service availability and resiliency capabilities, as well as the ability to accommodate change. Ø Availability Outages and performance incidents impact the business. Maintaining an executive-level view of network and application performance, as well as being able to rapidly troubleshoot incidents as they arise, are essential to seamless delivery. Ø Flexibility Growth proliferates changes across an IT environment, which stresses capacity limitations. Changes can come from new users, growth in data, or new applications. Supporting growth and change are essential to seamless delivery. Ø Resiliency Data protection, disaster recovery, and business continuity are essential to seamless hybrid cloud delivery, but they can be costly, difficult and unreliable. Hybrid Cloud Availability Technology Spotlight: Riverbed Web Content Optimization Ø Web page complexity. Stingray Aptimizer automatically decreases the round trips between browser and server, reducing page load times dramatically. Ø Web page size. Stingray Aptimizer automatically applies a number of smart reduction and compression algorithms that reduce page load by optimizing each page request based on web platform, browser type, and page structure. Ø Repeat visits. Stingray Aptimizer applies advanced techniques to ensure repeat visitors content is always kept up to date without rechecking or downloading the content again. When content changes on the server, the browser is automatically refreshed. Ø Browser diversity. Stingray Aptimizer automatically recognizes browser types and optimizes page resources without impacting the functionality of the page. Any number of factors can cause network performance problems, including device failure, application changes or Internet service problems, but the network team often has a hard time discovering a problem early and finding its source, due to lack of suitable tools. Improving hybrid cloud availability requires a combination of architectural robustness, effective monitoring, and prompt remediation. Ø Architectural robustness: Organizations can take advantage of hybrid cloud cost efficiencies by deploying an application across multiple public and private cloud data centers. Stingray Traffic Manager global and local load balancing solutions can then balance application traffic across these multiple cloud deployments, reducing risk and improving the performance and capacity of applications. Ø Effective monitoring and reporting: To deliver the hybrid cloud performance that the business needs, network and application teams need complete visibility into the applications, servers, and users on their network. Importantly, applicationawareness gives administrators the business context they need to monitor critical services and provide executive-level reporting. Cascade fulfills this need with network monitoring, network traffic recording and analysis, packet capture, and performance management capabilities that collect traffic data on the network, calculate performance metrics and alert administrators to problems before end-users are impacted. With executive-level reporting capabilities, Cascade provides an SLA dashboard for hybrid cloud deployments. Ø Rapid remediation: In complex hybrid cloud environments, incidents may be unavoidable, so when they occur, it s imperative that administrators have the tools to rapidly identify, diagnose, and resolve the issue. To accelerate troubleshooting and enable end-to-end network visibility, Riverbed consolidates all network data into a single data set with a top-down view, so administrators can navigate seamlessly from high-level application status down to detailed performance data to resolve an issue. 2011 Riverbed Technology. All rights reserved. 5

Hybrid Cloud Flexibility Environments are not static: as new applications and new users are introduced, existing network infrastructure can become stressed. With cloud solutions that fit a wide variety of environments, simplifying infrastructure, maximizing performance and flexibility, Riverbed customers can adapt their architecture as new applications are introduced, cloud offerings evolve, or business needs change. Two common causes for change in a hybrid cloud environment are rapid data growth and the development and deployment of new applications. Ø Data growth: The rapid growth in enterprise data adds cost and complexity to hybrid cloud infrastructure, particularly in meeting backup windows and recovery point objectives (RPO). Steelhead data streamlining can reduce bandwidth consumption by 65-95x, reducing the costs associated with data growth. Meanwhile, traditional means of data protection such as tape backup, duplication and vaulting are struggling to keep up with the twin demands of data growth and flat budgets. The Whitewater cloud storage gateway applies network and storage deduplication to backup and archive workloads to reduce their storage footprint by 10-30x on average and enabling storage in the cloud, which offers a much lower TCO than traditional backup methods. Ø New applications: New applications can strain existing network connections, but Ø WAN Optimization (WOC) with Riverbed QoS technology, administrators gain control to prioritize new and existing applications according to the business requirements. Another challenge in meeting the demands of continuous delivery is the need to develop and test new applications in production-like environments. As a software Application Delivery Controller (ADC), Stingray Traffic Manager can be cost-effectively used in test, development and production, ensuring that new applications are built and deployed in consistent environments. Stingray Traffic Manager can be installed in all three environments, and after cloning in-production applications for developer use, users can be seamlessly migrated to upgraded versions. Hybrid Cloud Resiliency Disaster recovery and business continuity remain top IT priorities, as the cost of downtime increases. Meanwhile, business evolves increasingly global and nonstop operations, increasing the complexity of protecting systems. Regardless of where data and applications reside, Riverbed has solutions to streamline data protection, improve disaster recovery preparedness, and ensure business continuity in the event of an outage or disaster. Ø Data protection: The first step in hybrid cloud resiliency is ensuring that data is being backed up from one data center to secondary location. Data protection operations can be streamlined by reducing the bandwidth requirements for storage replication between public and private clouds by up to 95%. Data is replicated not only significantly faster, but contention for network resources with other applications can be controlled with Steelhead s QoS technology. Riverbed Whitewater also helps customers adopt cloud storage for backup and archive by integrating with their current data protection tools, and accelerating, encrypting, and deduplicating data as it is transferred to the cloud provider for long-term safe keeping. Finally, data protection processes can be monitored with Riverbed Cascade, using behavioral analysis to detect and alert IT administrators to any anomalies in backup procedures. Ingredients for Seamless Hybrid Cloud Deliver: Ø Application Delivery Controller (ADC) Ø Network Performance Management (NPM) Ø Cloud Storage Gateway (CSG) Ø Disaster recovery: The second step in hybrid cloud resiliency is ensuring that data and operations can be restored in the event of a site-wide outage with a minimal amount of data and economic opportunity lost. By streamlined data protection and enabling off-site storage at secondary or public cloud 2011 Riverbed Technology. All rights reserved. 6

data centers, Riverbed solutions can improve a company s disaster recovery capabilities. Faster replication means more frequent backups, improving recovery point objectives (RPO); and faster restoration, whether from secondary private data centers or public cloud storage, means significant improvements to recovery time objectives (RTO). Ø Business continuity: The final step in hybrid cloud resiliency is enabling operations to continue seamlessly. Stingray technologies can route internal and external users away from unavailable or failed servers, ensuring application remain availability. With Steelhead Mobile, employees can remain productive in the event of facilities closure, with fast access to applications and files, from home or temporary locations. Conclusion Hybrid clouds give organizations the freedom to use an evolving mix of public and private cloud resources in pursuit of increasing profits or productivity, reducing costs, and managing risk. Enabling a high-performance hybrid cloud requires several considerations in each of the planning, migrating, and ongoing delivery phases: Discovery and dependency mapping is the key to IT planning Major IT projects such as IT consolidation and cloud deployments have one thing in common: the need to discover what s running on the network, and who is using it. The discovery phase of IT projects has historically been time-consuming, expensive and error-prone, with several months spent scanning the network, finding dependencies and updating asset lists. Regardless of whether the plan is consolidating a few servers to public cloud or moving entire data centers to a private cloud model, Riverbed can help accelerate the project, achieve planned savings sooner and reduce the risk of outages and failed cutovers. Within a few hours, Riverbed can discover all systems, identify server and client dependencies and locations, and show the information in a graphical, interactive format. Optimization enables hybrid cloud migration, performance, and scalability Riverbed hybrid cloud solutions enable organizations to accelerate the movement of resources between private data centers to and from various cloud service providers. Delivering the enhanced speed and efficiency required to effectively deliver high performance services spanning private and public IT infrastructures requires optimizing network, web content, and application delivery. Acceleration improves productivity, reduces required capacity and costs, and enables component systems to be deployed where desired independent of distance. Ensure ongoing service delivery by focusing on availability, flexibility, and resiliency Whether delivering services from private cloud data centers, or enabling business units to leverage the public cloud, IT is becoming the internal service provider for organizations. Riverbed hybrid cloud solutions offer significant improvements to IT service availability, data protection and disaster recovery capabilities. Service level agreements, recovery point and recovery time objectives can be more easily monitored and met. At the same time, analytics and troubleshooting are simplified, backup/restore and replication is accelerated, and cloud storage can be encrypted and deduplicated. Riverbed hybrid cloud solutions enable customers to find the optimal balance of resources hosted in their own data centers or various cloud service providers. They can build an architecture that discovers and places applications, servers, or storage where each meets the specific requirements, and still retain the ability to smoothly adapt and evolve their environment in the future. About Riverbed Riverbed delivers performance for the globally connected enterprise. With Riverbed, enterprises can successfully and intelligently implement strategic initiatives such as virtualization, consolidation, cloud computing, and disaster recovery without fear of compromising performance. By giving enterprises the platform they need to understand, optimize and consolidate their IT, Riverbed helps enterprises to build a fast, fluid and dynamic IT architecture that aligns with the business needs of the organization. Additional information about Riverbed (NASDAQ: RVBD) is available at www.riverbed.com. Riverbed Technology, Inc. 199 Fremont Street San Francisco, CA 94105 Tel: (415) 247-8800 www.riverbed.com Riverbed Technology Ltd. One Thames Valley Wokingham Road, Level 2 Bracknell. RG42 1NG United Kingdom Tel: +44 1344 31 7100 Riverbed Technology Pte. Ltd. 391A Orchard Road #22-06/10 Ngee Ann City Tower A Singapore 238873 Tel: +65 6508-7400 Riverbed Technology K.K. Shiba-Koen Plaza Building 9F 3-6-9, Shiba, Minato-ku Tokyo, Japan 105-0014 Tel: +81 3 5419 1990 2011 Riverbed Technology. All rights reserved. 7