12.0.0 (2026-08-31)

Bug Fixes

• close the microservice client when the ping fails (8826bc6)
• http: create a dedicated HttpService instead of resolving from DI (163bfef), closes #2667
• microservice: stop RMQ pingCheck from asserting a queue per probe (9afc532), closes #2680
• mongoose: ping the database instead of checking readyState (e1c6486), closes #2661
• set shutting_down before awaiting the graceful shutdown timeout (1615c29)

Features

• ship the package as ESM (fae6663)
• add attempt() function (ef9ab99)
• add degraded status for health indicator (1cfb8af)
• add responseTime to the HealthCheckResult (a0434ae)
• @nestjs/terminus no longer has runtime dependencies

BREAKING CHANGES

• the package is now ESM-only and requires Node ^20.19.0 || ^22.12.0 || >=24.0.0.
• the deprecated HealthIndicator base class and its getStatus() have been
  removed. Migrate custom indicators to HealthIndicatorService:
  super.getStatus(key, isHealthy, data) becomes
  this.healthIndicatorService.check(key).up(data) / .down(data).
• the deprecated error classes are no longer exported: HealthCheckError,
  ConnectionNotFoundError, TimeoutError, StorageExceededError,
  UnhealthyResponseCodeError, MongoConnectionError.
• throwing inside a health indicator no longer produces a 503. In v11, a thrown
  HealthCheckError was unwrapped into the health check result; now any thrown
  error is treated as an unexpected failure (500) and only a returned down
  result marks the check unhealthy (503).