Add info about limits of using EventDB for metrics (#476)

manototh · web-flow · commit 1328cc43805f · 2025-12-09T21:22:27.000+01:00
diff --git a/query-data/metrics/overview.mdx b/query-data/metrics/overview.mdx
@@ -32,14 +32,15 @@ Logs and metrics are different forms of event data. It’s often possible to rep
 - **Standardized data types**: OpenTelemetry defines a small set of metric types with clear semantics. The metric type tells you how to query the data—no guessing how to aggregate or calculate rates.
 - **Faster queries**: Metrics data is stored in a format optimized for producing time series and other popular metrics visualizations. No need to process the data at query time.
 
-### When to use EventDB for metrics
+### When to use MetricsDB or EventDB for metrics
 
-If you can instrument your architecture to use the OTel Collector, you should use MetricsDB for storing metrics data. If you can’t use the OTel Collector, there are some circumstances when you can still use EventDB for metrics-style workloads:
+| Use MetricsDB when | Use EventDB when |
+|--------------|--------------|
+| You can send metrics using the OTel Collector or another OTel-compatible SDK | You can’t use OTel-compatible tooling |
+| High-frequency measurements (many per second) | Low-frequency point-in-time gauges |
+| You need rate calculations or percentiles | Simple aggregations are sufficient |
 
-- **Your metrics are gauges**: Point-in-time measurements like temperature or memory usage that don’t need rate calculations.
-- **Query complexity is low**: You’re querying small quantities of data, and can afford to derive metrics at query time.
-
-EventDB provides flexible querying through APL without requiring structured metric types.
+When in doubt, use MetricsDB. It’s designed for metrics workloads and scales better as volume grows.
 
 <Prerequisites />
 
diff --git a/query-data/metrics/query-metrics.mdx b/query-data/metrics/query-metrics.mdx
@@ -118,5 +118,5 @@ Available aggregation functions:
 
 For example:
 - To calculate the number of series, use `group using count`.
-To sum the values of all series, use `group using sum`.
+- To sum the values of all series, use `group using sum`.
 - To group data by the `project` and `namespace` tags using the `sum` aggregation, use `group by project, namespace using sum`.
