Prevent Problems Before Thresholds Are Crossed
Other performance and capacity management solutions focus on collecting, monitoring, and alerting when individual metrics cross a pre-established threshold level. While this is helpful in identifying bottlenecks which either are, or soon will be, affecting performance, it does not provide IT Operators with the necessary actions to alleviate the issue. In a virtualized environment, with shared resources, transitory workloads, and high degrees of variability, determining the right action after the fact is a challenging task. Even in environments with only a few hundred VMs, several thousand metrics, configuration constraints, and resulting outcomes must be considered in the decision-making process. Given this complexity, it is not possible for even the most experienced administrator to solve these problems – and a mathematical approach is required to continuously tune and optimize the environment in a healthy state.
More importantly, threshold-based management is both subjective and reactive. Determining the right thresholds is impossible in a dynamic environment and “learning what’s normal” just propagates bad practices and suppresses alerts during periods of known contention. Rather than focus on threshold-based management, the emphasis should be on ensuring that the resources required by your applications in general, and your mission-critical ones in particular, are always available. To do this you must have a management solution that:
- Has real time insight into the performance of the applications and the underlying infrastructure
- Is continuously aware of dynamic changes across the environment in order to prevent problems before they occur
- Possesses the ability to intelligently tune (in the form of tangible actions) the configuration across the environment to optimize workload performance and system utilization
What sets VMTurbo apart is that it looks at the entire environment holistically through its unique architecture and Economic Scheduling Engine. The technology consistently profiles how capacity in the physical infrastructure is being used and the nature of workload and application demands – providing you with the right intelligence to maintain health, reduce contention and avoid problems from occurring in the first place.