Documentation
¶
Overview ¶
Package postgres implements distributedlock.Locker against PostgreSQL session- scoped advisory locks (pg_try_advisory_lock). It uses an existing platform/database.Client for connection management.
IMPORTANT — TTL semantics: PostgreSQL advisory locks have no native TTL. The TTL argument to Acquire/Refresh on this provider is ADVISORY ONLY. The lock is held until Release is called or the dedicated session is closed (e.g. by a network failure or by Locker.Close). Callers that need a hard upper bound on lock duration should impose it via context deadlines or by tracking elapsed time themselves; the Refresh method on this provider only verifies that the underlying session is still alive — it does not extend any expiry on the database side.
Each Acquire reserves a dedicated *sql.Conn from the database client's pool so that the matching pg_advisory_unlock targets the same session. The Locker tracks outstanding connections internally and releases all of them on Close.
Index ¶
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
func NewPostgresLocker ¶
func NewPostgresLocker( cfg *Config, db database.Client, logger logging.Logger, tracerProvider tracing.TracerProvider, metricsProvider metrics.Provider, cb circuitbreaking.CircuitBreaker, ) (distributedlock.Locker, error)
NewPostgresLocker constructs a new postgres-backed distributedlock.Locker.
Types ¶
type Config ¶
type Config struct {
Namespace int32 `env:"NAMESPACE" envDefault:"0" json:"namespace"`
}
Config configures a Postgres-backed distributed locker. Namespace is mixed into the lock-id hash so independent services that share a Postgres cluster do not collide on the same advisory-lock id space.
Source Files