New Adaptive Logs Feature Lets Teams Delete Low-Value Logs Before Ingestion
Grafana Cloud today announced the public preview of custom drop rules for its Adaptive Logs service, enabling platform and observability teams to eliminate wasteful log lines before they are ever written to storage.
“For years, teams have struggled with logs that offer no insight—health checks, forgotten DEBUG output, or verbose INFO from rarely-used services,” said Jane Doe, Senior Product Manager at Grafana Labs. “Our new drop rules give them direct control to remove that noise, reducing costs and operational overhead instantly.”
How Drop Rules Work
Drop rules allow users to define logic based on log labels, detected log levels, or line content. Logs matching a rule are dropped before being written to Grafana Cloud Logs, ensuring only valuable data is retained.
“This is a game-changer for centralized teams who want to enforce cost-saving policies without touching each service’s configuration,” added Doe. “A single rule can drop all health-check logs across the entire organization.”
Three Ways to Use Drop Rules
- Drop by level – Eliminate noisy DEBUG or INFO logs that eat the logging budget.
- Sample chatty logs – Apply a drop percentage to retain a representative sample while discarding the rest.
- Target specific producers – Use label selectors combined with log level or text to isolate high-volume, low-value services.
Background: The Log Noise Problem
Observability teams typically ingest gigabytes of log data daily, much of which never provides useful signals. Health-check pings, repetitive debug output, and verbose informational messages inflate storage costs and slow down queries.
Until now, the only way to remove such logs involved complex configuration changes at the application level, requiring coordination across many teams. Grafana Cloud’s new drop rules bypass that burden.
What This Means
Drop rules are part of a complete cost-management system within Adaptive Logs. The system evaluates logs in this order: exemptions (preserve critical logs), drop rules (eliminate known noise), and patterns (apply optimization recommendations).
“Teams can now combine intelligent recommendations with their own expertise to achieve maximum savings,” said Doe. “We’re giving them the tools to turn off the faucet, not just mop the floor.”
The feature is available in public preview starting today. Early adopters report immediate reductions in log volume and associated costs, with some teams cutting their monthly log spend by more than 40%.
For full documentation, visit the Adaptive Logs drop rules guide.