Key Considerations for Leveraging Virtualization and Keeping Your Applications Available
Key Considerations for Leveraging Virtualization and Keeping Your Applications Available
By Don Norbeck published: Monday, July 30 2007
By Don Norbeck, director, product development, SunGard Availability Services
Introduction
Today’s IT organizations are faced with the daunting task of optimizing all aspects of their departments, including people, processes and technology. Virtualization—one particularly exciting example of this trend—has been defined as the pooling of various IT resources in a way that masks the physical nature and boundaries of those resources from the users. It promises to unlock underutilized server capacity through partitioning, enabling the ability to run multiple operating systems on one server. The most obvious benefits of virtualization are not just the ability to achieve higher server and overall computing infrastructure utilization, but also improved responsiveness and flexibility—since virtual resources can be moved or modified dynamically to reflect changing business needs.
Disaster recovery represents an innovative application of virtualization technology. Traditionally, disaster recovery has been achieved through the use of redundant infrastructures and systems housed in remote locations, away from primary datacenters. While an effective strategy, many IT organizations find it difficult if not impossible to buy and maintain “two of everything.”
Through the virtual partitioning of hardware to support multiple operating systems, virtualization purports to do away with the “two of everything” approach by delivering a level of redundancy that can be used for disaster recovery purposes—either on the same hardware or on a different piece of equipment within the same data center. Known as virtualized DR, this practice enables local availability within a single data center by facilitating failover from one partition to another partition in the same server, or in another server located in the same data center.
Virtualization is rapidly gaining momentum as a means of supporting business critical applications. However, the notion of local availability delivered on a single server or within a single datacenter directly contradicts a key tenet of the Seven Tiers of Disaster Recovery, which states unequivocally that a DR strategy with no off-site data may mean, quite possibly, no recovery (the Seven Tiers were originally defined by SHARE, a user group for IBM mainframe computers, in the early 1990’s to help identify the various methods of recovering mission-critical computer systems as required to support business continuity. Since then, the Seven Tiers have evolved into an industry-standard discussion).
Neither virtualization nor virtualized DR is a panacea. Rather, they are complex initiatives requiring careful, ongoing consideration of key issues, including:
Increased criticality of virtual servers
Increased complexity for server administration and management
Change management and architecting the infrastructure migration from physical to virtual environment
Balancing investment versus higher utilization rates
Need for new skill sets and processes
Increased Criticality of Virtual Servers
One of the first assumptions of a virtualization implementation is that it will allow an organization to cut down on the number of servers it needs to support its applications. This may be true, but while the number of physical servers may decrease, each one—by virtue of its supporting multiple applications—becomes significantly more critical, which in turn requires close consideration of disaster recovery strategies to support each virtual machine.
Because they are more critical, virtual machines may be more vulnerable to individual “choke points.” Case in point: one of the first issues often overlooked when contemplating a virtualization solution is the high power consumption and high heat output of a single server hosting multiple virtualized servers. The reason for this is that virtualized servers cause the hardware CPUs to run at a higher utilization rate, requiring greater power and resulting in greater heat output, which in turn requires advanced (and often expensive) cooling mechanisms. A single malfunction of a cooling mechanism could be a recipe for a crash with costly consequences. In addition, virtualized servers require power redundancy (or the ability to withstand a single circuit failure). Without it, one accidental trip over a plug could bring down dozens of applications.
Increased complexity for server administration and management
Many organizations view server virtualization as a way to lower costs, and believe reducing staff levels and lowering server support ratios will be part of that. But recent research shows that this is not always the case, with initial installations of server virtualization products actually increasing the need for support labor because the environment is now more complex.
Virtualization often drives improved total cost of ownership and better use of hardware resources. However, in looking at potential staff savings, hardware support issues like installation and configuration represent just a small percentage of support staff time, which also includes software installation, configuration management, patching, capacity and performance monitoring, problem analysis and resolution, likely making labor reduction minimal.
Virtualization can serve the double effect of making these routine software-related tasks more complex, requiring higher investments of time and attention from IT staff. The end result may be a distracted and stretched IT staff that fails to pay due attention to important disaster recovery considerations.
Change management & architecting the infrastructure migration from physical to virtual environment
Virtualization places greater emphasis on a comprehensive understanding of interdependencies—or, how a change to a particular hardware element may impact another infrastructure element and the applications and operating system instances that element supports. With virtualization, the task of taking and maintaining an inventory of servers, their application workload roles or identities (including systems and application software tied to each specific server model) becomes more difficult and critical, while the interdependencies multiply and become more complex.
In a virtualized environment, there is a higher likelihood of failure to recognize interdependencies—and as a result, a higher risk for unplanned downtime which can effect multiple applications and become much more costly. Disaster recovery strategies can mitigate these risks by backing up and ensuring availability for virtual machines. As one respondent in an SQL Magazine survey noted, “Interdependence oversights fall into the ‘oops, I forgot’ class of problems.”[1]
Patch management provides an example. In a non-virtualized environment, an organization may have to plan for application downtime around a Microsoft® Windows® Update patch being installed on the host server. In a virtualized environment, patching and rebooting a single host server creates a much more significant impact, because numerous operating systems and applications may be running on the server.
Virtualization also requires IT administrators to understand what applications are co-dependent and which must reside on the same server. For example, if application A is dependent on data inputs from application B, the two applications need to reside on the same server. If they are housed on separate servers, downtime for the server hosting application B may result in data discrepancies between the two applications, resulting in business inefficiencies or worse, costly errors.
The constantly changing nature of business applications and the IT infrastructures supporting them also requires the regular capture and storage of virtualized IT blueprints, also known as images. Images are often set up in-house and undocumented, and the challenge of re-staging an image can be so great that in the event of a disaster, an IT department may have to start from scratch—thereby hurting recovery point and recovery time objectives.
Balancing investment versus higher utilization rates
Virtualization often requires significant investments in the form of one-time, up-front expenses. Organizations must consider and understand the true cost scope in order to determine if higher server utilization rates (and the corresponding promise of lower costs) merit the level of investment and potential for higher operational risk. Organizations that forge ahead must evaluate how to address these needs in the most cost-efficient manner while upholding the need for enhanced disaster recovery protection.
Up-front costs can include:
Prototyping: IT departments often begin virtualization projects by running production-like pilots to uncover any technical glitches or previously unknown requirements. Some server vendors may lend a box for free, or at a significantly reduced price, but limit the deal to the piloting stage of the project. Start-up help from a vendor’s professional services may not be included in the pre-sale.
Preparing for a server consolidation negotiation: Before enlisting a professional consulting firm to assist with server consolidation, organizations must comprehensively audit their existing systems and applications. These internal audits can be both time-consuming and expensive, but organizations that fail to carry them out risk being unable to clearly indicate requirements and expectations to vendors.
Understanding the nature of your applications: Virtualization is not appropriate for every type of application, and organizations should have a strong grasp of their existing and future application needs before investing heavily in virtualization. For example, applications that are CPU or I/O intensive are not ideal for virtualization, because of the need for fully dedicated resources and unshared high capacity.
Software licensing: Virtualization may result in drastic changes to software pricing schemes, with software vendors moving away from more traditional CPU pricing models to usage-based utility pricing. Organizations have benefited from CPU pricing, since it is relatively straightforward and allows employees, customers and partners to use software freely, without careful monitoring. The move toward “usage-metering” could mean escalating prices.
Datacenter costs: Virtualization can require a certain amount of datacenter fortification, and the associated costs can quickly add up. For example, virtual servers are heavier than their non-virtualized counterparts. Many datacenters were simply not designed to support this weight—especially considering that the average datacenter is 17 years old.[2] Virtualized servers also require advanced (and often expensive) cooling mechanisms and greater power supplies.
Need for new skill sets and processes
Virtualization calls for new IT capabilities including enabling applications to dynamically acquire more resources to accommodate peak performance; CPU capacity prioritization; and storage I/O and network traffic prioritization. This represents a significantly higher level of strategic planning than is customary for many IT departments, and a whole new level of skill sets that may not be inherent in existing staffs, no matter the size.
Another important skill set to consider is security. Virtualization requires a different information security skill set, because in virtualized environments the security must be provided at the hypervisor level versus the virtual machine level. Compounding the problem is the fact that virtualization, as with any emerging technology, may be a prime target of new security threats.
Prior to a virtualization or virtualized DR deployment, organizations need to implement comprehensive training programs. Virtualization software vendors may provide training on their products, but it may not encompass all of the knowledge areas required including hardware, firmware, middleware and software.
Conclusion
Fortunately, in spite of these risks, virtualization experts are available to help organizations identify opportunities for virtualization as part of larger disaster recovery strategies, and link virtualization initiatives to disaster recovery planning—ultimately helping to mitigate risk while optimizing infrastructure performance.
For example, today’s virtualization experts are:
Helping organizations develop and implement virtualization strategies through staff training, assistance with IT audits, prototyping and testing, strategy development and finally, change management and roll-out.
Taking a leadership position in developing tools and techniques to support virtualized disaster recovery. One example is automated provisioning, which will enables the rapid development and deployment of organizations’ virtual images on remote systems. Another example is virtual image capture and storage. For organizations supporting virtualized environments on their own premises, virtual image snapshots can be securely and reliably stored on remote systems.
Delivering virtualized environments as managed services, which are supported by the industry’s most advanced datacenter technologies and diverse range of hardware platform expertise;
Offering datacenter moves and consolidations–-while maintaining operational continuity throughout. Many organizations leverage datacenter moves and consolidations as opportunities to simultaneously implement virtualization, and virtualization experts make it possible to concurrently maximize these initiatives.
If implemented and managed properly, virtual machines and virtualized DR infrastructures can dramatically shrink recovery point and recovery time objectives. However, if not properly managed, virtualization can introduce significant availability risks which can hamper business resiliency and overall competitive position. By comprehensively addressing the issues raised in this article, organizations can position themselves to maximize virtualization initiatives, reap the full benefits of virtualized DR and ultimately protect and strengthen their businesses.
_______________________________________
REFERENCES
1. Brian Moran, SQL Server Magazine. “Top Causes for SQL Server Downtime.” SQL Server Magazine. 28 August 2003.
2. IDG Research. “Business Continuity: 2007 Research Brief.” 17 July 2007.
The above information is presented as general information only. SunGard Availability Services urges you to become and stay informed of virtualization developments and issues by consulting other reliable sources.
SunGard and the SunGard logo are trademarks or registered trademarks of SunGard Data Systems Inc. or its subsidiaries in the U.S. and other countries. All other trade names are trademarks or registered trademarks of their respective holders.
