What changed
Sentry’s deadline for legacy Alerts API migration passed on August 17, 2026. Deprecated endpoints now return HTTP 410 and Sentry says there are no backwards-compatible endpoints or temporary exemptions. The replacement architecture splits detection from routing: Monitors define what to watch and create issues when conditions are met; Alerts define who gets notified and can be connected to multiple monitors. Sentry’s migration guide also points Terraform users to the newer resources introduced in provider version 0.15.4.
Why it matters
This is not a cosmetic API rename. Integrations that create or manage issue and metric alerts programmatically can now fail outright, and the replacement model changes how builders should represent alerting configuration. Detection conditions, issue creation and notification routing are separate resources, which can reduce duplication but requires migration work in internal tooling, Terraform modules and scripts. Teams that relied on old endpoints need to treat 410 responses as a broken production control path, not a transient outage.
The old endpoints are gone, not merely deprecated
Sentry’s migration guide says users had until August 17 to move, after which the legacy endpoints become unavailable. Deprecated calls return 410 Gone. Sentry explicitly says there is no compatibility endpoint or temporary exemption, so retrying or waiting will not restore an integration.
Monitors detect; Alerts route
The new architecture separates the condition that creates an issue from the action that notifies people. Metric, Cron, Uptime and other monitor types live under Monitors, while Alerts attach notification logic to issues and can connect one routing rule to multiple monitors. This is more composable than the old one-rule-per-alert shape but changes object models and IDs for API clients.
Existing UI rules were migrated, custom automation was not
Sentry says existing metric alert rules were migrated automatically in the product UI, with thresholds becoming Metric Monitors and notification actions becoming connected Alerts. That does not remove the need to update external code that still calls legacy endpoints. Internal provisioning systems, scripts and Terraform modules need to use the replacement APIs and resource types.
Terraform has a defined replacement path
Sentry points Terraform users to new provider resources, including `sentry_alert` and `sentry_metric_monitor`, added in version 0.15.4. Older `sentry_issue_alert` and `sentry_metric_alert` resources are also slated for deprecation in a future major provider version, so infrastructure-as-code users should migrate before a provider upgrade forces a second break.