Docs / Monitoring / Insights Page and Trend Analysis Usage

Insights Page and Trend Analysis Usage

Published · Last updated: May 2026 · 4 min read

Who this is for

Users who want to understand infrastructure trends over time — not just current status — to make proactive decisions about scaling, optimization, and capacity planning.

What you will complete

Navigate the Insights page, interpret trend data, and use it to identify servers that are approaching capacity or underutilized.

Before you begin

  • Servers must have been running for at least 7 days for meaningful trend data.
  • Navigate to Insights in the left sidebar.

What the Insights page shows

The Insights page is different from the real-time Monitoring dashboard. While Monitoring shows current metric values and recent history, Insights aggregates data over longer periods to surface patterns and anomalies.

Insights includes:

  • Resource trend charts — CPU, RAM, and disk trends per server over the past 7 to 30 days
  • Capacity alerts — servers approaching resource limits based on trend projections
  • Underutilization flags — servers that are consistently below 10–15% CPU/RAM (potential candidates for downsizing)
  • Anomaly highlights — periods where resource usage deviated significantly from the server's historical baseline

Reading trend charts

Each server on the Insights page shows a trend line for each metric. Healthy patterns to look for:

Stable (healthy): The line is relatively flat with predictable daily cycles (higher during business hours, lower at night). This is the expected pattern for most web application servers.

Gradual upward trend (attention needed): Disk usage climbing steadily over weeks is common and normal — but will eventually require action. Insight's capacity projection shows when you will run out at the current rate.

Increasing peaks (investigate): If daily CPU peaks are getting higher week over week with no corresponding increase in traffic, this may indicate a memory leak or a growing background process.

Capacity projections

For disk usage specifically, Insights provides a capacity projection: "At the current growth rate, this server will reach 90% disk usage in approximately X days."

Use this to plan ahead:

  • If a server is projected to fill in 7 days, schedule a disk expansion or cleanup now.
  • If a server is projected to fill in 60 days, add it to your monthly maintenance checklist.

Underutilization detection

Insights flags servers that are consistently using less than a threshold of their provisioned resources. This is a cost optimization signal.

What to do with underutilized servers:

  • Consider downsizing the server to a smaller, cheaper instance type.
  • Check whether the server is actually needed, or can be combined with another server.
  • Ask AI Pilot: "Server-name shows as underutilized. Is it safe to downsize it and what should I check first?"

Using Insights for capacity planning

Monthly review checklist using Insights:

  1. Open Insights and sort servers by disk growth rate — highest first.
  2. For any server projected to hit 90% disk within 30 days, plan a disk expansion or cleanup.
  3. Sort by average CPU — identify consistently high-CPU servers that may need more compute.
  4. Identify underutilized servers and evaluate whether downsizing makes sense.
  5. Check the anomaly highlights — any server with an unusual usage pattern in the past 7 days.

What success looks like

  • Insights data appears for servers that have been running at least 7 days.
  • Capacity projections show meaningful estimates (not "N/A") for disk usage.
  • You can identify at least one server that needs attention before it becomes a problem.

Common errors and fixes

"Insights shows 'No data' for my servers" Cause: Servers have been running for fewer than 7 days, or the monitoring agent is not installed. Fix: Ensure the monitoring agent is active on all servers (see KB-06-10). Return to Insights after 7 days of data collection.

"Capacity projection shows an unrealistic date (like 200 years)" Cause: The server's disk usage is nearly flat — it is growing so slowly the projection is essentially infinite. Fix: This is actually good news. No action needed.

"I see a server flagged as underutilized but I know it handles burst traffic" Cause: Insights uses average utilization — burst traffic may not show up in average metrics. Fix: Switch to the real-time Monitoring dashboard and use the 7-day chart to see if there are regular traffic spikes. Insights underutilization flags are a starting point, not a definitive recommendation.

Related articles