Grafana Cloud Unveils Adaptive Logs Drop Rules: Instantly Slash Log Noise and Costs

Breaking News – Grafana Cloud has launched a public preview of Adaptive Logs drop rules, enabling platform teams to eliminate wasteful log lines before they are ingested, reducing noise and cutting costs immediately.

“With drop rules, teams can now define custom logic to drop low-value logs—like health checks or debug messages—without touching infrastructure code,” said a Grafana product manager. “It’s a direct way to stop paying for log noise.”

Background

Most observability teams deal with logs they know are pure noise: throwaway health checks, forgotten DEBUG statements, or verbose INFO from rarely used services. These logs inflate storage and processing bills without providing insight.

“The hard part was always getting rid of them without changing every application’s configuration,” noted a senior observability engineer. “Now there’s a simple, centralized tool.”

How Drop Rules Work

Drop rules are evaluated in priority order when a log line arrives in Grafana Cloud. They are one of three mechanisms in Adaptive Logs:

• Exemptions – Protected logs pass untouched.
• Drop rules – Custom rules apply a drop percentage or 100% drop.
• Patterns – Optimization recommendations handle remaining logs.

Users can create rules based on log labels, detected log levels, or line content. Examples include dropping all DEBUG logs, sampling repetitive logs by a percentage, or targeting a specific noisy service with a label selector.

What This Means

This update means teams no longer need complex configuration changes to remove unwanted logs. A single rule can enforce standards across all services, saving money and reducing storage load.

“Drop rules complement our existing intelligent recommendations, giving teams full control over log cost management,” a Grafana engineer explained. “It’s a complete system: exemptions protect critical data, drop rules eliminate known noise, and patterns handle the rest.”

For more details, visit the official Adaptive Logs documentation.