CA and VMware today announced the results of a study the companies sponsored investigating the impact of virtualization on data center operations.
The study, conducted by the IT Process Institute (ITPI) in December 2008, identifies specific procedures and controls that should be considered to reduce risk as organizations virtualize business-critical systems and when production virtualization objectives beyond server consolidation evolve to high availability, disaster recovery, and dynamic resource management scenarios.
ITPI collected data from 323 North American IT organizations about their server virtualization practices. The survey focused on the procedures and controls used most frequently to manage the technology. Based on an analysis of the procedural changes the IT organizations made to optimize the benefits and reduce the risks of virtualizing production data centers, the IT Process Institute developed the following recommendations for each level of maturity:
- Baseline Maturity Practices: For those organizations consolidating servers and virtualizing business-critical systems in production environments, ITPI identified 11 practices, which include host access, configuration and provisioning controls, virtual machine provisioning, and capacity and performance management.
- High Maturity Practices: For those organizations expanding beyond server consolidation to high availability and disaster recovery objectives in an otherwise static environment, ITPI recommended 25 practices. The practices help IT organizations quickly respond to performance-impacting conditions with a high degree of configuration standardization, provisioning with approved build images, and using a “trust but verify” strategy for change process and configuration compliance.
- Dynamic Computing Practices: For those organizations pursuing dynamic resource management objectives, ITPI recommended 12 practices including controls primarily in the area of configuration discovery, change approval and tracking; capacity and performance management; and overall process maturity needed to support automation.
Analysis of hard performance measures revealed a statistically significant correlation between the use of recommended practices and various hard outcome measures, such as:
- The use of host access controls predicts a higher level of availability measured by minutes of downtime per month.
- The use of capacity management practices predicts better service support performance with measures such as the rate of incidents resolved within service level agreement limits, and mean time to repair large outages.
- The use of provisioning automation and discovery practices predicts a higher rate of production systems that match target configuration.
Leave a Reply