Query optimization
Diagnose slow queries using the slow query log, EXPLAIN, and Performance Schema. Rewrite expensive patterns and design indexes around real production access paths, not assumptions.
MySQL Consulting
I help engineering teams improve MySQL performance, review database architecture, reduce operational friction, and connect database work to the cloud and application layers around it — whether you are running RDS MySQL, Cloud SQL, or self-hosted MySQL 5.7 or 8.x.
Diagnose slow queries using the slow query log, EXPLAIN, and Performance Schema. Rewrite expensive patterns and design indexes around real production access paths, not assumptions.
Improve connection handling, InnoDB buffer pool configuration, monitoring, backup discipline, and deployment safety so the database stops being a recurring source of incidents.
Support MySQL 5.7 to 8.x upgrades, topology changes, and schema redesigns when the current setup is constraining delivery or growth. Includes RDS and Cloud SQL migration paths.
Most MySQL performance work starts with the slow query log — enabling it if it is off, or actually reading it if it has been collecting data nobody has looked at. From there, EXPLAIN and EXPLAIN ANALYZE show the execution plan the optimizer chose and why it may be wrong. Performance Schema fills in the gaps: lock waits, connection saturation, handler statistics that reveal access patterns the query itself does not make obvious.
On RDS MySQL and Cloud SQL, the tools are the same but the constraints are different — you cannot tune the kernel, parameter groups have limits, and some configurations require a restart with a maintenance window. Understanding what is and is not possible in a managed environment changes the diagnosis and the fix.
InnoDB buffer pool sizing, connection pooling behavior, and the interaction between query performance and application-layer caching are often where the real leverage is — not just individual query rewrites.
Result
Query response time reduced from 3.8s to 95ms on a high-traffic RDS MySQL instance.
After enabling and analyzing the slow query log, the bottleneck was a query hitting a large orders table with no usable index on the most selective column. A covering index was added, the query was rewritten to avoid a filesort, and InnoDB buffer pool sizing was adjusted to keep the hot data in memory. Average response time dropped from 3.8 seconds to under 100ms with no schema migration required.
A fixed-scope Database Performance Audit is $1,500 and takes one week — slow query log review, execution plan analysis, index assessment, and a prioritized fix list. Larger implementation engagements are scoped per project. See the full offer details.
Yes. Most production MySQL work I do runs on RDS MySQL or Cloud SQL for MySQL. The diagnostic approach is the same as self-hosted MySQL, but managed environments have specific constraints around parameter group changes, storage autoscaling, and maintenance windows that affect how fixes get implemented.
Yes. MySQL 8 introduced breaking changes in collation defaults, GROUP BY behavior, and authentication that catch teams off guard if they run the upgrade without a compatibility review first. I can audit your queries and schema for 8.x incompatibilities, flag what will break, and help sequence the migration safely.
Slow query log review, EXPLAIN and EXPLAIN ANALYZE for the highest-impact queries, index coverage and access pattern assessment, connection and InnoDB configuration review, and a written report with a prioritized list of changes and the expected impact of each. Delivered within one week.
Related services