As a cost center, IT Departments are frequently charged with reducing bottom line costs and improving efficiency, while facing the challenge of continuing to meet business objectives and competitive SLAs. When thoughtfully planned out and executed, server consolidation can be the answer to help improve efficiency and reduce infrastructure operating costs.
What is Server Consolidation: Server Consolidation is a process of condensing applications, processes, management and servers across fewer and/or virtual devices and locations.
Why Consolidate?
Realize a Significant Return on Investment: Consolidating servers creates a more efficient use of hardware and can reduce the number of operating system software licenses and resources needed to manage the environment, thus reducing operating costs and potentially increasing system utilization. These actions create power savings, increase efficiency and can reduce your carbon footprint, all of which can be very easy to justify financially.
If executed thoughtfully, consolidation almost always results in a positive return on investment. You can realize both immediate and long-term software license and hardware costs reductions that far outweigh the cost of the time and effort spent to plan and implement your consolidation plan.
However, without a proper plan in place, you can make investments in inefficient or incompatible technology, make mistakes that can lead to performance degradation, catastrophic system failures, and end up spending costly hours on repairs.
Whether this is your organization’s first attempt or 21st, we’ve laid out a few tips to help guide you through the process.
Step One: Analyze the Environment
As a first step in planning a server consolidation effort, you should begin by monitoring capacity, performance, and workload to determine a baseline of simple workload patterns.
Utilize monitoring tools over a representative period of time and measure the impact of the current workload on: CPU utilization, RAM utilization, and I/O rate.
Determine busiest hours, and analyze any spikes. Make sure you are acting on fresh analysis, and not dated information.
Step Two: Choose a Consolidation Method
Common consolidation processes include:
- Logical Consolidation is used when implementing common processes and management across a range of applications.
- Physical Consolidation is used when condensing servers across fewer locations.
- Rationalized Consolidation is used when implementing multiple applications on fewer, more powerful platforms.
- Virtualized Consolidation is used when partitioning one physical server into multiple virtual servers.
The Best Approach: Consolidate and Virtualize
Is it possible to consolidate what you already own? Using your analysis, identify any underutilized servers and begin to plan which applications you can centralize, based on workload and operating system compatibility.
Then, determine where virtualization can be implemented. Virtualization is the process of utilizing virtualization software to partition one physical server into multiple virtual servers, and provides the best and most cost-effective method for consolidation.
According to a recent review of virtualization trends, technology writer Stephen J. Bigelow says, “The future holds promise for server virtualization and consolidation. Server technologies continue to advance, offering more processor cores and memory for the same amount of capital investment. This means that each technology refresh can potentially host more VMs and further reduce the total number of physical servers.”
The cost of operating multiple physical servers (which includes OS licensing, support, and maintenance costs) greatly outweighs the cost of any virtualization software licensing.
When choosing virtualization software, you want to make sure that the product is compatible with application data storage and failover needs, and the hardware you’ll be using. And, don’t assume that you can reuse any existing servers. You may need to account for the cost of new hardware with the resources and memory designed specifically for virtualization, and factor that cost into the project. Servers that do not have the capabilities to be utilized for virtualization can be repurposed for other tasks, such as disaster recovery.
Step Three: Construct your Strategy and Present Your Plan
To keep your project at a high ROI, you want to ensure that the time and testing needed to migrate data and applications does not cancel out any cost savings realized by consolidation. Use the 80/20 rule to decide which servers and applications to consolidate: identify and rate candidates for consolidation by application complexity. The more complex an application, the greater time it will take to test. The least complex apps are better candidates, requiring less testing effort (the 80%).
Generally, database servers make the easiest candidates for consolidation, since client server applications can easily be pointed to a different database server.
You may not want to attempt more complex application moves until you have a baseline of the time and effort migration requires. Realize that there are some candidates (the 20%) that will not offer a good enough return to include in your consolidation effort, due to the complexity of the migration and testing effort.
Once analysis is complete and all consolidation approach, goals and objectives have been identified, present your plan to upper management and end users. Emphasize the return on investment: don’t forget to factor in licensing and support cost savings, and potential efficiency gains.
Step Four: Configure and Migrate Data
After a solid executive plan and any new hardware and software are in place, it is finally time to migrate the data and applications.
Planning the migration of data is the most crucial part of the procedure. The strategy above identified the best candidates for consolidation. Now, following the strategy drafted in step above, you need to lay out a rolling schedule of migration, testing, and outage windows designed to impact the least amount of users.
Coordinate closely with vendors and application users to conduct testing, and plan multiple short outage windows, for the least impact on your customer base.
Step Five: Take the Opportunity to Make Additional Improvements
Implement Important OS and Application Upgrades: Take advantage of the consolidation effort to upgrade OS and application software, for even more potential performance gains. Since you will already be planning outages and coordinating testing efforts, it’s the perfect time to incorporate any needed patching.
Implement Database Monitoring: Consider implementing a centralized monitoring service that simplifies database monitoring. As a baseline, ensure you automate the monitoring of disk and file space usage, database up checks, performance indicators, and missed backups. When you evaluate monitoring tools, look for services that consolidate the alerting process. New tools, such as Aspect Consulting’s Prodative® Watchdog EMS consolidate the monitoring and alerting process across your entire environment, regardless of platform. You can view alerts via a web-based portal, and all alerts are summarized in a single daily email. You want to ensure that your Admins are not inundated with a high number emails from each server, increasing the risk of missing critical issues that are buried too deeply in an inbox.
Step Six: Document You New Configuration
As a final step, make the effort to document the new configuration. Accurate documentation is crucial, and will help increase the supportability of the environment and improve troubleshooting efforts. As a minimum, document server configuration information (include server names, rack space, locations), access information, and back and recovery information.
Conducting a final study in the immediate weeks after the consolidation is also a good idea – assess what went right and what went wrong. This documentation will prove invaluable in future troubleshooting and future consolidation efforts.
Posted by: Sharon Steenkamer/ Sandy Snyder
