Skip to content

RDS MariaDB Read Replica FreeableMemory really low after minor version upgrade

0

We recently upgraded our MariaDB instance from 10.6.17 to 10.6.20 and noticed a sudden drop in FreeableMemory. The memory has not recovered and remains around 150 MB (instance type: t3.medium) even after a couple of days.

1 Answer
0

The issue you're experiencing with low FreeableMemory after upgrading your RDS MariaDB instance is concerning, but there are a few things to consider and potential steps to take:

  1. Version-specific issues: While your specific version upgrade (10.6.17 to 10.6.20) isn't mentioned in the available information, it's worth noting that version upgrades can sometimes introduce unexpected behavior or bugs related to memory management.

  2. Memory allocation: The drop in FreeableMemory could be due to changes in how the new version allocates memory for various database operations or caches. For example, the InnoDB buffer pool or dictionary cache might be using more memory in the new version.

  3. Monitoring and investigation: Use CloudWatch and Performance Insights to closely monitor your database's performance metrics. Pay attention to metrics like swap usage, CPU utilization, and database connections.

  4. Configuration review: Check your DB parameter group settings, especially those related to memory management like innodb_buffer_pool_size. Ensure they are appropriate for your instance type and workload.

  5. Workload changes: Consider if there have been any changes in your application's workload that coincided with the upgrade, which might be contributing to increased memory usage.

  6. Optimization: Review your queries and database schema for potential optimizations. Inefficient queries or excessive use of temporary tables can impact memory usage.

  7. Instance sizing: If the issue persists, you may need to consider upgrading to a larger instance type with more memory.

  8. Replication impact: If this is a read replica, ensure that replication is functioning correctly, as replication issues can sometimes lead to unexpected resource consumption.

If the problem continues without improvement, you may want to consider opening a support case with AWS to investigate further or potentially rolling back to the previous version if feasible. Additionally, always ensure you're running the latest minor version, as subsequent releases often include bug fixes for issues like this.

Remember to test any configuration changes or upgrades in a non-production environment first to avoid potential impacts on your production workload.
Sources
MySQL RDS' FreebleMemory goes down while taking DB Dump | AWS re:Post
Troubleshooting a MariaDB read replica problem - Amazon Relational Database Service
Troubleshooting memory usage issues for Aurora MySQL databases - Amazon Aurora

answered 8 months ago

You are not logged in. Log in to post an answer.

A good answer clearly answers the question and provides constructive feedback and encourages professional growth in the question asker.