Our Performance Optimization Process
Our performance optimization methodology
Our performance optimization methodology has been designed to produce predictable optimization results for our clients. By performing an end to end review of all relevant server performance metrics, our comprehensive Pre-Optimization Healthcheck begins the process with identifying all high-value areas of potential performance improvement.
After the Pre-Optimization Healthcheck, the optimization begins. Our structured and analytical scenario-based optimization approach ensures that numerous solutions are exhausted and thoroughly stress-tested before the recommended optimization is provided.
Scenario-based performance optimization
Traditional approaches to performance tuning are highly iterative trial and error process that heavily rely on the knowledge of a single expert. This process often involves a repetitive approach of testing one individual performance change after another in sequence.
Our scenario-based optimization approach combines years of experience in performance management with a highly robust, automated set of diagnostic and testing tools to deliver consistent, jaw-dropping performance improvements.
At the heart of our approach is a highly sophisticated process of simultaneously testing multiple scenarios in every problem area for performance improvement. We then combine the results of these scenarios to come up with an optimal performance solution.
Significant code redesigns have not only have expensive development impact, but also hidden costs associated with required long QA cycles. While we recognize that some systems may have the flexibility for rapid changes, in our performance tuning approach, we strive to minimize code changes and to limit modifications to the existing system architecture.
Areas of optimization
- Severe performance degradation and system crashes
- Slow and resource-intensive queries
- Long-running locks and deadlocks
- Long-running batch jobs
- High-CPU utilization on the server
- Poor cache-hits ratio and memory pressure
- Sub-optimal scalability patterns
- Inefficient use of parallel processing
- Transaction rollbacks and failures
Final Results: Delivering predictable and benchmarked performance improvements
We provide predictable and benchmarked SQL Server performance improvements. At the end of our process, we deliver to you a complete set of scripts and associated documentation for all proposed changes and a benchmark report that quantifies the individual and combined impact of these recommended optimizations on current performance and future scalability.