Log noise is a persistent challenge for platform and observability teams. Health check pings, forgotten DEBUG statements, or verbose INFO logs from rarely used services inflate storage costs and obscure critical signals. Centralized teams have long struggled to eliminate this waste without cumbersome infrastructure changes. Enter Adaptive Logs drop rules—a new feature in Grafana Cloud (now in public preview) that lets you define custom rules to drop low-value logs before they're ever written to Cloud Logs. This approach reduces noise and saves money immediately, complementing intelligent optimization recommendations already available in Adaptive Metrics and Adaptive Traces. Here are 10 things you need to know.
1. What Are Adaptive Logs Drop Rules?
Drop rules are a mechanism within Adaptive Logs that allow you to create logic using any combination of log labels, detected log levels, or line content. When a log line arrives in Grafana Cloud, it's evaluated against your rules before storage. If a rule matches, the log is either dropped entirely or sampled at a specified percentage. This gives you fine-grained control over which logs are worth keeping. Unlike traditional filtering that requires changes to application code or logging libraries, drop rules are managed centrally in the cloud, making them a non-invasive solution for reducing log volume.
2. How Drop Rules Directly Reduce Costs
Every log line ingested into Grafana Cloud Logs incurs a cost—whether it's useful or not. Drop rules cut costs by preventing worthless logs from ever being written. For example, a single rule with a 100% drop rate can eliminate all health check logs from every service in your environment. This not only lowers your monthly bill but also frees up storage and query performance for high-value data. The savings are immediate and can be tracked via the Adaptive Logs dashboard, where you see volume reductions in real time.
3. Target Logs by Level: Drop DEBUG and Low-Level INFO
Many teams enable verbose logging during development but forget to turn it off in production. Drop rules make it easy to target specific log levels. For instance, you can create a rule that drops all DEBUG logs or filters out INFO logs from services that produce too many. Simply specify the level in the rule criteria. This is one of the quickest wins for reducing log volume because DEBUG and verbose INFO logs often represent a large portion of ingested data while offering minimal operational value.
4. Sample Chatty, Repetitive Logs Without Losing Context
Not all noisy logs should be dropped entirely. Some repetitive logs—like periodic status updates from a batch job—are useful for occasional troubleshooting but overwhelming at full volume. Drop rules let you apply a drop percentage, effectively sampling the stream. You can keep, say, 10% of those logs to maintain a representative sample. This preserves visibility into the service's behavior while slashing the total volume. The sampling is applied deterministically, so you won't miss critical error spikes.
5. Zero In on a Specific Noisy Producer
Sometimes a single service starts emitting an unexpectedly high volume of low-value logs—maybe after a misconfigured update. Drop rules allow you to target that service by specifying a label selector (e.g., service=my-chatty-app). You can combine this with additional criteria like log level or a text string. This precision means you don't have to affect other services that are behaving well. It's a surgical approach to log management that empowers platform teams to act without bothering service owners.
6. Combine Multiple Criteria for Complex Filtering
Real-world log noise often comes from a combination of sources. Drop rules support multiple conditions—label selectors, log levels, and text patterns—in a single rule. For example, you could drop all logs where the service is my-app, the level is INFO, and the line contains the word "heartbeat." This granularity ensures you only filter out exactly what you intend, avoiding accidental removal of valuable logs. Rules are evaluated in priority order, so you can create a hierarchy of filters.
7. Understand the Order: Exemptions, Drop Rules, Then Patterns
Drop rules are one component of a complete volume management system. When a log line arrives, it goes through three stages in order:
- Exemptions: Logs that match an exemption pass through untouched—no sampling or dropping.
- Drop rules: Next, the first matching drop rule applies its drop rate (from 0% to 100%).
- Patterns: Finally, remaining logs may be subject to optimization recommendations (patterns) for further volume reduction.
This pipeline ensures that critical logs are never dropped, while known noise is efficiently removed upfront.
8. Drop Rules Complement Adaptive Metrics and Traces
If you're already using Adaptive Metrics or Adaptive Traces in Grafana Cloud, you'll be familiar with the concept of custom drop rules. The same capability is now extended to logs. This unified approach means you can apply consistent data-management policies across all telemetry types. For example, you might drop DEBUG logs from a service while simultaneously downsampling its traces and filtering its metrics—all with similar rule structures.
9. Practical Use Cases: Health Checks, Batch Jobs, and More
Drop rules shine in common scenarios:
- Health check logs: Drop 100% of all health-check-endpoint logs across every service.
- Batch processing jobs: Sample repetitive log output from cron jobs or ETL pipelines with a 90% drop rate.
- Libraries with noisy logging: Target third-party library logs that only produce noise.
- Migration periods: Temporarily reduce volume from a service being refactored, without touching its config.
Each use case can be implemented with a single rule, making it easy for platform teams to enforce standards without developer friction.
10. Getting Started with Drop Rules Today
Drop rules are now in public preview for Grafana Cloud users. To start, navigate to the Adaptive Logs section in your Grafana Cloud instance and click on Drop Rules. You can create rules using the UI or via the API. The documentation provides step-by-step guidance and examples. Best of all, you can see the impact immediately in the volume reduction charts. Start with one rule—like dropping DEBUG logs—and watch your costs drop. There's no risk: you can disable or modify rules at any time.
Conclusion: Adaptive Logs drop rules give platform teams a simple, powerful way to eliminate log noise without infrastructure changes. By combining targeted dropping, sampling, and compatibility with existing exemptions and patterns, you can reduce costs, improve signal-to-noise ratio, and focus on what matters. Start using drop rules today and take control of your log volume.