Flaky test: TestServer_HealthMonitoring_Enabled

[JAORMX]

TestServer_HealthMonitoring_Enabled in pkg/vmcp/server/health_monitoring_test.go is flaky due to tight timing thresholds.

The test uses:

• check_interval=50ms
• unhealthy_threshold=1
• timeout=10ms
• degraded_threshold=5ms

Under load or with the race detector enabled, the health check can exceed the 5ms degraded threshold, causing the test to see "degraded" instead of the expected "healthy" status.

health_monitoring_test.go:152:
    Error:      Not equal:
                expected: "healthy"
                actual  : "degraded"

The fix should increase the timing thresholds to be more tolerant of CI/local machine variance.