New Sciple is live. Connect your AWS in 10 minutes and explore the platform with a sandbox workspace. Book a demo
sciple

Monitoring & Metrics Explorer

The metrics explorer is where you query and chart time-series data. It speaks to two backends, Prometheus and Amazon CloudWatch, and presents both through the same chart, so you can investigate from one place.

Before you start. The explorer reads from your connected metrics providers. If a tab is empty, connect a source first. See Connecting data sources.

Querying Prometheus

On the Prometheus tab, write a PromQL expression and run it. Pick the time window and the sampling step to control how much detail the chart shows, then read the series in the chart and the summary table below it.

The Prometheus metrics explorer with a PromQL expression and a multi-series memory chart.
  1. 1 Switch between Prometheus and CloudWatch.
  2. 2 Enter a PromQL expression.
  3. 3 Choose the time range and step.
  4. 4 Click Run Query, and each matching series is drawn as its own line, with the latest, mean, max, and min summarised underneath.

Querying CloudWatch

The CloudWatch tab builds a query without you writing any expression. Choose the AWS data source, then pick a namespace, a metric within it, and a statistic. Add dimensions to narrow the result to a specific resource. The metric catalogue loads from the connected account, so you only see metrics that actually exist.

The CloudWatch metric builder with namespace, metric, statistic, and dimensions controls.
  1. 1 Pick a Namespace (for example AWS/EC2 or AWS/RDS).
  2. 2 Choose a Metric from that namespace.
  3. 3 Select a Statistic such as Average, Sum, Maximum, Minimum, or a percentile.
  4. 4 Add Dimensions to target a specific resource.
  5. 5 Click Run Query to chart the result.
Tip. Found a query worth keeping? Add it to a board as a metric panel. See Creating dashboards.

Reading the chart

Every series is coloured and labelled so overlapping lines stay legible. The window you pick at the top applies to the whole chart; widen it to spot trends, narrow it to inspect a spike. The step controls resolution, so a smaller step shows more points but takes longer to query over long windows.

Related