Fix Magento MySQL CPU Spike
Budget: $10 – $30 CAD
Our Magento 2.4.7-p5 shop is running on AlmaLinux under Plesk and the MySQL process sits at 500–600 % CPU around the clock. Slow queries pour into the log even after the indexing and query-level tweaks I have already applied, so the issue is clearly deeper than basic schema adjustments.
Access & stack
• SSH and Plesk access on a dedicated VPS
• Exact MySQL build still needs confirmation, but it is the stock package shipped with the current AlmaLinux repo.
• Redis and Varnish are available but disabled while troubleshooting to remove variables.
What I need
1. A deep diagnostic session that pinpoints the queries, engine settings, or Magento modules driving the sustained load.
2. Concrete tuning steps applied directly in MySQL (InnoDB parameters, buffer sizes, thread concurrency, etc.) and, if required, Magento configuration or extension clean-up.
3. A brief report that shows before/after metrics so I can track the improvement.
Acceptance criteria
• Average CPU for mysqld stays below 80 % during normal traffic peaks.
• Slow query log trimmed to sporadic entries (<10/min).
• No regressions in storefront or admin functionality.
If you are comfortable profiling Magento on busy stores, diving into EXPLAIN plans, and leaving the system measurably faster, let’s get this solved today.
Access & stack
• SSH and Plesk access on a dedicated VPS
• Exact MySQL build still needs confirmation, but it is the stock package shipped with the current AlmaLinux repo.
• Redis and Varnish are available but disabled while troubleshooting to remove variables.
What I need
1. A deep diagnostic session that pinpoints the queries, engine settings, or Magento modules driving the sustained load.
2. Concrete tuning steps applied directly in MySQL (InnoDB parameters, buffer sizes, thread concurrency, etc.) and, if required, Magento configuration or extension clean-up.
3. A brief report that shows before/after metrics so I can track the improvement.
Acceptance criteria
• Average CPU for mysqld stays below 80 % during normal traffic peaks.
• Slow query log trimmed to sporadic entries (<10/min).
• No regressions in storefront or admin functionality.
If you are comfortable profiling Magento on busy stores, diving into EXPLAIN plans, and leaving the system measurably faster, let’s get this solved today.