Change Overlays is in beta.
Overview
As teams iterate, deploy code, and continually make changes to their applications and services, it can be difficult to find the exact change that caused a spike in errors, an increase in latency, or slower page load times. Use Change overlays to identify when a recent change is causing performance issues within your application or services and find the source of the problem.
View the moment a change occurs in the context of your Datadog observability data to pinpoint issues to specific releases, correlate changes with metrics, and troubleshoot faster. Change overlays supports APM service deployments.
Overlay changes on graphs
To get started, click Show Overlays in the upper right corner of your dashboard.
Overlays automatically appear on timeseries graphs filtered with the service
tag for services configured with version
tags. To enable deployments in your APM services, add version tags to your configuration.
Click on any event overlay to open a side panel with more information and analyze the impact of your change.
Show faulty deploys
Use the toggle in the overlays panel to only show faulty deployments that could be impacting your metrics.
Override automatic detection
Override the automatic service detection by using the search bar to find the service of interest.
Analyze the impact of your change
Click on any overlay on your graph to open a change analysis page, allowing you to understand the status and impact of you change.
APM deployments
For APM deployments, you can:
- Compare the version selected to overall service performance for requests, errors, or latency
- View the rollout of your version by
region
, env
, or datacenter
- See new Error Tracking issues that have been introduced with the new deployment
- Check related infrastructure that your service is running on
FAQ
When do overlays appear?
For APM deployments, overlays appear on timeseries graphs that:
- Are filtered by the
service
tag in the query - Have the
service
set up with the version
tag
What are deployments scoped to?
For APM deployments, an env
must be specified. If you have an env
or datacenter
template variable set in your dashboard, deployments are filtered to match the selection. Otherwise, the env
defaults to prod
.
Further Reading
Additional helpful documentation, links, and articles: