Spam Filters

Overview

Spam filters in Dash0 help you control costs by preventing the storage of excessive or unnecessary telemetry data. These filters allow you to specify rules that drop spans, logs, and metric data points during ingestion, ensuring you don't get charged for telemetry that provides little value to your observability goals. Rather than collecting everything indiscriminately, spam filters let you focus on the most relevant data for troubleshooting and monitoring your systems.

How Spam Filters Work

Spam filters operate at the ingestion pipeline level. When telemetry data matches your defined filter criteria, Dash0 automatically drops the matching data before it's stored, and you aren't charged for it. This process happens in real-time as data flows into Dash0.

The key benefits include:

• Reduced storage costs by eliminating low-value telemetry
• More focused and relevant data for analysis
• No charge for filtered data
• Dataset-specific configuration options

Configuring Spam Filters

You can configure spam filters directly within the tracing, logging, and metrics explorers:

The flag as spam button is located right within the logging, tracing and metric explorers.

1. Navigate to the tracing, logging, or metrics explorer
2. Define filter criteria for the data you want to exclude
3. Click the filter button (P) to create a spam filter
4. Confirm the filter creation in the dialog

Once configured, the filter will become active within a few minutes. Note that any data dropped cannot be recovered, so create filters carefully.

Clicking on the flag as spam button brings up a dialog that asks for your confirmation.

Managing Spam Filters

All active spam filters appear in your dataset settings, where you can:

• Review existing filters
• Disable filters temporarily
• Delete filters that are no longer needed
• Add new filters

Different datasets can have different spam filter configurations, allowing you to apply more aggressive filtering in development environments while keeping production telemetry more complete.

Within the dataset configuration dialog you can access a list of all configured spam filters.

Filter Limitations

There are some important considerations when using spam filters:

1. Filters operate on individual telemetry items (spans, logs, or metrics), not on aggregated data
2. You can only filter based on attributes present in the transmitted data
3. Attributes that Dash0 automatically correlates (like those via resource centricity) are unavailable for filtering
4. When filtering spans, you might end up with incomplete traces
5. It can take up to 5 minutes for a new filter to become active
6. Dropped data cannot be recovered

Export Filters to OpenTelemetry Collector

For users who want to implement filtering earlier in their telemetry pipeline, Dash0 allows exporting spam filters as OpenTelemetry Collector configurations. This uses the filterprocessorCopy and OTTL (OpenTelemetry Transformation Language) to create compatible filter configurations that can be deployed in your infrastructure.

From the spam filter, you can directly jump to an OpenTelemetry Collector configuration in OTelBin. Your configured filter is replaced by an OTTL expression.