Review and apply recommendations
Adaptive Logs provides you with a list of patterns and recommendations on which logs to drop based on usage patterns. You can filter, review, and edit drop rates to suit your needs.
After reviewing, you apply the drop rates to implement the changes. The logs are then dropped upon ingestion into Grafana Cloud, reducing unnecessary log storage costs while ensuring that your observability remains intact.
Adaptive Logs recalculates its pattern list and recommended drop rates every 24 hours using a rolling 15-day window of your log ingest and query behavior. When a pattern is first detected, Adaptive Logs delays displaying it for 7 days to gather enough usage information to recommend a reasonable drop rate.
- Navigate to Adaptive Telemetry > Adaptive Logs.
Select the Patterns tab to review patterns and recommendations for drop rates.
On the Patterns tab, you can:
- Filter patterns by service and frequency, making it easier to decide which log lines to drop
- View the query frequency, volume, drop rate, recommended rate, and projected savings for each pattern
- The Recommended column shows the recommended percentage of log lines to drop
- The Current(%) column shows the percentage of lines currently being dropped
- Expand a pattern to view a breakdown per service, including service name, percentage, savings, and volume
- Click the Open in Explore icon to see log lines matching a pattern
Apply all recommendations
Applying all recommendations is the most effective and efficient way to optimize your log management. It saves time compared to manually reviewing each recommendation individually and gives you immediate cost savings.
Click Apply all recommendations from the Adaptive Logs user interface.
The Apply recommended drop rate(s) dialog box provides information about ingest volume vs projected volume and the projected savings of applying the recommendations.
Review the recommended drop rates in the dialog.
Click Apply to confirm.
In the next milestone, you’ll learn how to create exemptions to protect critical log streams from being dropped.
More to explore (optional)
At this point in your journey, you can explore the following paths: