Documentation
¶
Index ¶
- type HealthCheck
- type ManagedSubprocess
- type SubprocessManager
- func (m *SubprocessManager) BeginShutdown()
- func (m *SubprocessManager) CrashChan() <-chan error
- func (m *SubprocessManager) ForwardSignal(sig os.Signal)
- func (m *SubprocessManager) Shutdown(timeout time.Duration)
- func (m *SubprocessManager) Start(name string, cmd *exec.Cmd) error
- func (m *SubprocessManager) WaitHealthy(ctx context.Context, name string, check HealthCheck) error
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
This section is empty.
Types ¶
type HealthCheck ¶
type HealthCheck struct {
URL string // e.g. "https://127.0.0.1:4444/health/alive"
Interval time.Duration // polling interval
Timeout time.Duration // overall timeout before giving up
TLS *tls.Config // optional TLS config for HTTPS health endpoints
}
HealthCheck defines how to check if a subprocess is healthy.
type ManagedSubprocess ¶
type ManagedSubprocess struct {
Name string
Cmd *exec.Cmd
// contains filtered or unexported fields
}
ManagedSubprocess tracks a running subprocess.
type SubprocessManager ¶
type SubprocessManager struct {
// contains filtered or unexported fields
}
SubprocessManager starts, monitors, and shuts down a set of subprocesses. Fail-fast: if any subprocess exits unexpectedly, the crash is reported via CrashChan. Once Shutdown begins, subsequent subprocess exits are expected and are NOT reported as crashes. Shutdown is reverse order of start.
func NewSubprocessManager ¶
func NewSubprocessManager(log *logger.Logger) *SubprocessManager
NewSubprocessManager creates a new subprocess manager.
func (*SubprocessManager) BeginShutdown ¶
func (m *SubprocessManager) BeginShutdown()
BeginShutdown marks the manager as shutting down, suppressing crash reporting for any subprocess that exits from this point forward. Call this as early as possible in the shutdown sequence — before stopping servers or other components that may cause cascading subprocess exits. Shutdown() also sets this flag, but calling BeginShutdown early closes the window between signal receipt and Shutdown() invocation.
func (*SubprocessManager) CrashChan ¶
func (m *SubprocessManager) CrashChan() <-chan error
CrashChan returns a channel that receives an error when any managed subprocess exits unexpectedly. The CP main loop selects on this alongside signal handling.
func (*SubprocessManager) ForwardSignal ¶
func (m *SubprocessManager) ForwardSignal(sig os.Signal)
ForwardSignal sends the given signal to all managed subprocesses.
func (*SubprocessManager) Shutdown ¶
func (m *SubprocessManager) Shutdown(timeout time.Duration)
Shutdown sends SIGTERM to all subprocesses in reverse start order, waits up to timeout for each to exit, then sends SIGKILL if needed. Sets the shuttingDown flag if BeginShutdown was not already called.
func (*SubprocessManager) Start ¶
func (m *SubprocessManager) Start(name string, cmd *exec.Cmd) error
Start launches a subprocess and begins monitoring its PID. Stdout/stderr are forwarded to the CP's stderr (visible in docker logs).
func (*SubprocessManager) WaitHealthy ¶
func (m *SubprocessManager) WaitHealthy(ctx context.Context, name string, check HealthCheck) error
WaitHealthy polls a health endpoint until it returns 200 or the timeout expires. Returns an error if the subprocess crashes before becoming healthy.