feat(graceful-shutdown): add enhanced production-ready shutdown sequence

[mag123c]

PR Checklist

Please check if your PR fulfills the following requirements:

• The commit message follows our guidelines: https://github.com/nestjs/terminus/blob/master/CONTRIBUTING.md
• Tests for the changes have been added (for bug fixes / features)
• Docs have been added / updated (for bug fixes / features)

PR Type

• Bugfix
• Feature
• Code style update (formatting, local variables)
• Refactoring (no functional changes, no api changes)
• Build related changes
• CI related changes
• Other... Please describe:

What is the current behavior?

Currently, the graceful shutdown mechanism doesn't provide proper support for production environments where readiness and liveness probes need to be handled separately. When a pod receives SIGTERM, there's no way to mark the readiness probe as unhealthy while keeping the liveness probe healthy, which can lead to:

• Load balancers continuing to route traffic to terminating pods
• Requests being lost during pod restarts
• No proper time for load balancer to detect and stop routing traffic

Issue Number: #2569

What is the new behavior?

This PR introduces an enhanced graceful shutdown mechanism with multi-stage shutdown process:

1. New configuration options:
   - enableEnhancedShutdown: Enables the enhanced shutdown sequence
   - beforeShutdownDelayMs: Time to wait after marking unhealthy (default: 15000ms)
2. Multi-stage shutdown sequence:
   - Stage 1: Mark readiness probe as unhealthy via isApplicationShuttingDown()
   - Stage 2: Wait for load balancer to stop routing traffic (configurable delay)
   - Stage 3: Process remaining requests for gracefulShutdownTimeoutMs
   - Stage 4: Terminate application
3. Public API additions:
   - Exported GracefulShutdownService for integration with health checks
   - Added isApplicationShuttingDown() method to check shutdown status

Example usage:

// Module configuration
TerminusModule.forRoot({
  enableEnhancedShutdown: true,
  beforeShutdownDelayMs: 15000,
  gracefulShutdownTimeoutMs: 30000,
})

// Health check integration
if (this.gracefulShutdownService.isApplicationShuttingDown()) {
  // Return unhealthy for readiness probe
}

Does this PR introduce a breaking change?

• Yes
• No

[mag123c]

ci e2e fail reason: #2672