SQL Doctor is one of the most popular products in the Idera portfolio and as such we get a variety of questions from users on how it works, where do the recommendations come from, etc. We picked out some of the more common questions that have been asked that perhaps others would like to know the answers to as well.
Q: How does SQL doctor extract performance information?
A: SQL doctor leverages performance counters, wait stats, DMV, DMF, WMI, SQL Profiler traces and standard SQL to gather the information needed to produce the recommendations.
Q: How does SQL doctor determine what to recommend?
A: SQL doctor utilizes an intelligent algorithm developed by our own SQL Server performance experts which leverages years of proven performance information and industry standard best practices from Microsoft.
Q: When SQL Doctor performs an analysis, how much will it affect my server while it pulls the information?
A: There will be some low CPU utilization during the analysis phase depending on which categories are selected to analyze. We suggest using some discretion for multiple category analyses during high utilization timeframes or schedule the analysis during off hours or maintenance windows.
Q: Does SQL doctor use Microsoft Best Practice Analyzer tests or should I run MSBPA as well?
A. SQL doctor is not a replacement for MSBPA, there is some degree of overlap between MSBPA and SQL doctor. The main difference is that SQL doctor is a performance analysis/recommendation tool that finds your most urgent SQL Server problems (query, network, memory, objects..etc) and offers detailed suggestions for resolving those problems. MSBPA is designed for administrators who want to determine the overall conformance of their SQL Server instances and topology with established best practices for system events, settings and login permissions.
Q. What is the difference between selecting “Development” or “Production” during the interview questions?
A. With regards to selecting “development” vs. “production” the same collection and analysis is performed. The only difference is that the recommendations generated are tailored to one environment or the other.
Q: Why are the recommendations provided for my test environment different from those from my production environment?
A. Each environment can be different, based on the hardware in place, the amount of data stored, the workload placed on that server, memory, CPU, network and many other factors. Those variables can generate significant differences in the recommendations. Performance data collected during the analysis phase represents SQL Server activity during that particular point-in-time for each test or production server. A test server’s performance recommendation may differ somewhat from a production server because each environment will have varying resources and workloads during the analysis phase.
These are answers to just some of the questions we get from users. Hope you found some of this informative. Are there any other questions you might have about SQL Doctor? Let us know in the product forum – we are here to help!