The virtualization services for database are trending because of its benefits of economy, efficiency and agility. Recent research on 500 IT professionals revealed that the majority of them(76%) virtualized the databases and / or are using a hybrid approach of virtual and physical databases to support their environment. After virtualization, however, an additional set of complexities need to be considered for maintaining high levels of reliability, scalability, and add value to business (for e.g., backup to cloud)
In the opinion of professionals, the five following best practices should be in the crosshairs of every database administrator (DBA), not only to ensure efficient progress of a virtualization project, but also to maintain peak performance after Virtualization:
1. Do not be afraid to virtualize a large database with high transactional load. Modern virtualization technology is reliable and robust, and can provide a performance very close to the performance of databases in no operating system servers.
2. Create a good working relationship with the virtualization administrator. If you cannot make changes to the resources of the virtual machines (VM), the DBA will lose the greatest benefit of virtualization – agility. Unlike changes made on one physical server, which is a manual process, with a few clicks you can add additional resources of CPU, RAM and storage to a virtualized database server.
3. Use shared metrics to improve collaboration between IT groups. Avoid confusion and accusations, creating a unified understanding between the database staff, developers and system administrators and virtualization. These teams can work collaboratively to maintain considerations as containment server resources and bottlenecks in I / O under control.
4. List the metrics for the physical host and VM capabilities with SQL query performance. VMs metrics can be misleading because the operating system does not report metrics on resources accurately when running in a VM. When it is contextualizedx metrics based on query performance, the impact caused by noisy neighbors issues, events and VM resource provision can be quickly identified.
5. Monitor the SQL queries of response time before and after virtualization. It is essential to have the right tools in use for response time visibility. These data show that the performance of the database has been maintained or improved, and when there are performance problems, the analysis enable their rapid identification and resolution.