Problem solve Get help with specific problems with your technologies, process and projects.

Server consolidation's final phase: VM monitoring and migration

Once you’ve gathered all your server consolidation project data, you’re ready to perform the virtual server migration. And don’t forget VM monitoring.

In your server consolidation project, the final phase -- virtual server migration -- will depend on the number of servers you start with and the number of new servers you intend to add to the environment.

Once you’ve consolidated and you have your virtual machines (VM), you must continue VM monitoring to ensure the server consolidation project’s success. Here are some general guidelines to follow when formulating a virtual server migration plan:

  • Divide the virtual server migration process into several manageable segments or waves, and allow time between each wave for the IT staff to deal with any problems.
  • The analysis phase of your server consolidation project may show that there are enough servers for the end-state environment. However, your virtual server migration plan must take into account the purchase of a sufficient number of “swing” servers to accommodate the first wave of migrations.
  • Plan the wave sequencing so that you can free up existing servers that are part of the end-state environment first. Use performance monitoring tools to measure and track application performance as the virtual server migration process proceeds. If performance problems exist once you consolidate, you may have to adjust computing resources that have been assigned to the troubled virtual machine (VM). You can also adjust the server consolidation project and shift the troubled VM to another server.
  • You don’t need to move application and database instances if the present server’s performance is adequate.
  • Import the details needed to finish each VM into the completed application/database instance review sheet. Then divide that sheet into separate pages for each wave in the virtual server migration process. Each sheet must include the OS and OS version, where the instances can be found, and whether license keys are needed.
  • Include all network changes needed to support planned migrations for each wave, particularly if you are unable to reassign current server IP addresses to new VMs.
  • For each virtual server migration wave, create a fallback plan in case something goes wrong.

Preparing for end-state operations: VM monitoring
Virtual server migration plans should also include end-state preparations for the entire data center and its staff. Data-center and VM monitoring will ensure that your project has succeeded.

Give each of the following factors in your server consolidation project adequate attention.

  • Staff training. All staff should be trained before the virtual server migration begins; senior staff members should be involved in the planning process.
  • Management tool selection, purchase and installment. The products you use will depend on the virtualization software that you have purchased. For example, organizations using Microsoft Hyper-V will probably use System Center Virtual Machine Manager; VMware users will likely choose vCenter.
  • Storage area network (SAN) capacity. Ensure that your SAN has enough storage capacity. If current servers are using the SAN, you may need to expand its capacity. Performance bottlenecks may occur when smaller servers with higher read/write activity place high demands on the SAN.
  • Data backup plan. Determine if you need to update backup procedures and tools.
  • Disaster recovery plan: Update your disaster recovery plan to take into account the virtualized environment.
  • Configuration management database. You may need to update the data center’s configuration management database (CMDB) to include the new VMs. After the virtual server migration, the CMDB needs to reflect relationships between VMs and servers. If your CMDB doesn’t have an underlying database design that recognizes these new relationships and attributes, your configuration management process will quickly unravel.
  • Containing VM sprawl. Build specific procedures into the data center’s operational process to avoid VM sprawl—the accumulation of unnecessary VMs created at the request of development teams or other groups. VM sprawl is wasteful in several ways, particularly in that it requires software licenses, which sit unused on machines. All requests for temporary VMs should include an expiration date. Once that date arrives, take steps to decommission the VM and remove all software.

VM monitoring is essential after each virtual server migration wave and throughout the first few weeks of operation. As you complete your server consolidation project, make sure that all servers have the proper mix of application or database instances. If problems occur during this process, one or more of the application/database instances should be moved to an underutilized VM to balance the load.

About the author:
Malcolm Hamer
is an IT consultant and director at Acumen Solutions Inc., a business and technology consulting firm. Prior to that, Hamer has 20+ years of experience working in technology for global telecommunications organizations, the banking sector and the securities industry. Hamer is co-author of Telecommunications: a Systems Approach, Telecommunications Systems and Telecommunications Users’ Handbook. He also writes articles and frequently presents on a range of IT topics.

Dig Deeper on Server consolidation and improved resource utilization

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.