Cloud SQL is an essential component for managing and scaling databases in Google Cloud, and cost optimization is crucial when using it efficiently. Here are some best practices to help you reduce Cloud SQL costs and optimize resource usage:

1. Idle Instances: Idle resources are one of the main reasons for wasted cloud spending. Cloud SQL's Active Assist feature detects idle instances and provides recommendations and insights on how to reduce costs. You can use Cloud Scheduler jobs to trigger Cloud Functions that will start and stop instances automatically, saving up to 75% of the cost to run an instance per week compared to continuous usage.
   ‍
2. Overprovisioned Instances: Developers sometimes provision unnecessarily large instances, leading to unnecessary spending. Active Assist detects instances that are unnecessarily large for a given workload and provides recommendations for resizing. It uses machine learning and Google's Cloud SQL telemetry data to identify instances with low peak utilization, ensuring they can still handle peak workloads after being resized.
3. Network Egress: Charges for network egress can occur even when both the function and the Cloud SQL instance are in the same region. To avoid these charges, add a private IP address for the SQL instance, a Serverless VPC Connector, and change the database connection code to use the private IP address explicitly.
   ‍
4. Long-term Commitments: Leverage commit use discounts for predictable workloads on a one-year or three-year basis to achieve significant cost savings. You can save 25% on on-demand pricing for a one-year commitment and 52% for a three-year commitment.
   ‍
5. Logs Expiration and Retention: Control the retention period of logs by setting the expiration time. If you need to maintain logs for an extended period due to data governance or regulatory requirements, set a longer retention time or export logs to cheaper long-term storage options such as Google Cloud Storage.
   ‍
6. Use Labels & Tags to Group Instances: Assigning labels & Tags to your instances can help you group them based on specific attributes, such as project, environment, or application. This allows you to filter instances by labels & tags, reducing the number of instances processed and ultimately reducing costs.
   ‍
7. Disable Log Exports for Non-Essential Logs: Ensure that you only export logs essential for your use case if you have enabled log exports to destinations such as BigQuery or Pub/Sub. Exporting non-essential logs increases costs without providing significant value.
   ‍
8. Use Log Views: Log Views allow you to create a customized view of your logs, displaying only the log entries that meet specific criteria. This can help reduce logging costs by filtering out unnecessary logs.

Remember that cost optimization is an ongoing process, and you should continually monitor and adjust your usage to ensure that you are getting the most value for your investment in Cloud SQL.