What We Do
Your time is valuable, let us analyze your data and provide suggestions quickly.
MySQL Server Tuning as an OFF cpu analysis of your global status and global variable data to determine likely Suggestions to improve performance of your host server.
There are more than 300 status metrics (activity counts) available on your host.
There are more than 400 system server variables that manage resource capacities.
As new versions are released, additional status and server variables will be introduced. Releases we have already analyzed include Community Edition, Percona Server, MariaDB, AWS Aurora, AWS RDS, MS Azure, Debian, Drupal, Google Cloud, Linux, MacOS, Magento ecommerce, NDB, NGX, RHEL (Red Hat), Ubuntu, WooCommerce, WordPress.
Our goal is to analyze your data and provide suggestions for your my.cnf or my.ini configuration to improve your associate/client/customer experience and reduce user wait times. Typically our suggestions will be e-mailed to you within 4 business days.
For example, on any BUSY system, we will typically suggest
thread_cache_size=100 # from nnn V8 suggests CAP of 100 to avoid OOM
Why? Because No One Ever has fun with Out Of Memory. Creating threads is an intense operation that will be minimized with this variable.
threads_created count - dramatic reduction expected, especially if you had thread_cache_size=less than 50.
slow_query_log=ON # from OFF - unless you risk filling your HD
If you can not leave log ON all the time, start with 10 minutes ON, then OFF Analyzing and correcting ONE a day will yield very positive results.
Appropriate indexes frequently will resolve your slow query problem.
Best practices for applying our suggestions include:
A. keep a dated copy of your my.cnf or my.ini prior to any change to allow quick restore of last used configuration in a \history folder. If your \history folder includes these filenames with leading ccyymmddhhmm-(name) you will have a historical tracking solution.
B. your research/understanding of the role a particular variable
and how it could affect operation of MySQL prior to making a change.
C. monitor results for 24 hours before applying your next change.
If you are in a crisis situation, it may be appropriate to change more than one variable to survive or achieve some predictable degree of stability with your server.