Stopping build failures before they block our releases

Build failures and slow CI pipelines remain key operational challenges for engineering teams. It’s common for DevOps engineers to lose release time troubleshooting issues hidden in long job queues, agent shortages, or failing build stages.

To address this, we built a dashboard tailored to make our build process more transparent. It brings together the most important stats like - master branch reliability, build durations, recent failures, and agent activity into a single unified view.

This dashboard was built in close collaboration with our Director of Engineering, Tim Wheeler, who wanted an easy, reliable way to monitor build health daily. His feedback shaped the layout and priorities here: everything from the quick-glance summary of master branch failures to the detailed breakdown of agent usage and build job queues.

With every sprint, this dashboard helps us:

• Identify build slowdowns early and fix them before they impact delivery timelines or developer productivity.
• Pinpoint problematic stages and focus our efforts where they’ll have the biggest impact.
• Monitor build times and agent load to keep builds running smoothly.
• Prioritize fixes by seeing which builds wait the longest when agents are busy.
  
  

Dashboard walkthrough

1. Master branch failure

This tile is a scalar visualization showing the most important metric - the number of build failures during a production release. It uses the Build Runs data stream and includes a day-over-day comparison.

2. Overall failure rate for all branches

Shows the overall failure rate across all branches from the Build Runs data stream. It helps us see how stable our builds are and where we can improve.

3. Recent builds (last10)

This tile lists the 10 most recent builds from the master branch. We can quickly see which ones failed. Each entry includes the pipeline ID and a direct link to Azure DevOps, making it easy to drill down into details, check failed stages, and investigate issues.

4. Build Duration metrics

This section includes two tiles focused on the master branch over the last 7 days:

• Average Build Duration – a scalar tile comparing the current week’s average build time with the previous week.
• Average Duration Trend – a line graph showing how build times fluctuate day to day.

Monitoring average build time helps us spot when something isn’t quite right, for example, if a build suddenly takes much longer even though it doesn’t fail. These insights help us catch stuck or slow-running jobs early and keep the pipeline running efficiently.

5. Build stage failures

This view highlights which build stages fail most often. While it’s not something we need to react to every day, it’s a great way to spot recurring problem areas.

For example, we’ve noticed that publishing test results & Install packages tends to fail the most. Keeping an eye on this helps us identify flaky or misconfigured steps and prioritize fixes where they’ll have the biggest impact.

6. Job queues and Agent usage

This part of the dashboard focuses on pipeline efficiency, helping us see how busy our agents are and whether jobs are getting stuck in long queues.

It shows when agent usage spikes or when jobs start piling up, which can signal capacity issues or bottlenecks in the build process.

We use these insights both to react quickly when builds slow down and to monitor long-term trends, so we can improve resource allocation and overall pipeline efficiency.

Take Control of Your Build Health

This dashboard gives us more than numbers, it gives us clarity. By bringing Azure Devops build data into SquaredUp, we’ve turned raw metrics into insight, helping us keep builds reliable, and deliver more efficiently.

Ready to build your own? Create a free account and connecting your Azure DevOps, and start optimizing your builds today.