How drift detection works
1
Baseline capture
When you first connect a provider, InfraAudit snapshots the configuration of every discovered resource. This becomes the initial baseline for each resource.
2
Scheduled scans
The drift scanner runs every 4 hours by default. Each run fetches the current state from the cloud provider API and compares it against the active baseline.
3
Findings created
Differences that exceed the detection threshold become drift findings, stored with a severity and a
detected status.4
Alerts fire
If alert rules are configured, drift findings at or above the severity threshold trigger notifications to your configured Slack channels, email addresses, or webhooks.
Drift types
Severity levels
Viewing drift findings
1
Open Drift Detection
In the sidebar, click Drift Detection. The list shows all active findings, newest first.
2
Filter the list
Use the filter bar to narrow by severity, type, provider, status, or resource type. Start with Critical and High to focus on the most urgent items.
3
Open a finding
Click a drift to see its full detail: what changed, on which resource, at what time, and a side-by-side JSON diff of the baseline versus current configuration.
Drift detail
The detail panel for each finding shows:- Summary — what changed, on which resource, and when
- Diff — a side-by-side JSON comparison of the baseline and current configuration
- Affected resource — a link to the resource detail panel
- Timeline — detected time and last updated time
- Recommendations — suggested remediation steps
Resolving a drift
From the detail panel, you have three options:IaC drift
InfraAudit also detects IaC drift — cases where the live resource configuration no longer matches the Terraform or CloudFormation template that originally defined it. Upload your IaC files under IaC in the sidebar. InfraAudit parses the files, identifies the resources they declare, and compares their expected state against the current live state.Managing baselines
A baseline is the “known good” snapshot that drift detection compares against. You can:- Capture a manual baseline from the resource detail panel — useful before a planned change
- Promote the current live state to a new baseline after an intentional change — this resets the comparison point so the change no longer appears as drift
CLI
Next steps
- Remediation — apply automated fixes with a rollback window
- Alerts — configure severity thresholds for notifications
- Recommendations — view AI-generated fix suggestions