feat: 更新定时任务

This commit is contained in:
tbphp
2025-07-07 21:03:51 +08:00
parent e461f1af5f
commit 5daf1a8112
3 changed files with 6 additions and 136 deletions

View File

@@ -44,9 +44,6 @@ func (s *KeyCronService) Start() {
s.wg.Add(1)
go s.leaderElectionLoop()
// processResults still needs to run independently as it handles results from the Pool
s.wg.Add(1)
go s.processResults()
}
// Stop stops the cron job.
@@ -66,21 +63,17 @@ func (s *KeyCronService) leaderElectionLoop() {
case <-s.stopChan:
return
default:
// Attempt to acquire the leader lock
isLeader, err := s.tryAcquireLock()
if err != nil {
logrus.Errorf("KeyCronService: Error trying to acquire leader lock: %v. Retrying in 1 minute.", err)
time.Sleep(1 * time.Minute) // Wait for a while before retrying on error
time.Sleep(1 * time.Minute)
continue
}
if isLeader {
// Successfully became the leader, start executing the cron job
s.runAsLeader()
} else {
// Failed to become the leader, enter standby mode
logrus.Debug("KeyCronService: Not the leader. Standing by.")
// Wait for a lock TTL duration before trying again to avoid frequent contention
time.Sleep(leaderLockTTL)
}
}
@@ -88,10 +81,7 @@ func (s *KeyCronService) leaderElectionLoop() {
}
// tryAcquireLock attempts to set a key in the store, effectively acquiring a lock.
// This relies on an atomic operation if the underlying store supports it (like Redis SET NX).
func (s *KeyCronService) tryAcquireLock() (bool, error) {
// A simple implementation for the generic store interface.
// The RedisStore implementation should use SET NX for atomicity.
exists, err := s.Store.Exists(leaderLockKey)
if err != nil {
return false, err
@@ -100,12 +90,9 @@ func (s *KeyCronService) tryAcquireLock() (bool, error) {
return false, nil // Lock is held by another node
}
// Attempt to set the lock. This is not atomic here but works in low-contention scenarios.
// The robustness relies on the underlying store's implementation.
lockValue := []byte(time.Now().String())
err = s.Store.Set(leaderLockKey, lockValue, leaderLockTTL)
if err != nil {
// It's possible the lock was acquired by another node between the Exists and Set calls
return false, err
}
@@ -113,10 +100,8 @@ func (s *KeyCronService) tryAcquireLock() (bool, error) {
return true, nil
}
// runAsLeader contains the original logic that should only be run by the leader node.
func (s *KeyCronService) runAsLeader() {
logrus.Info("KeyCronService: Running as leader.")
// Defer releasing the lock
defer func() {
if err := s.Store.Delete(leaderLockKey); err != nil {
logrus.Errorf("KeyCronService: Failed to release leader lock: %v", err)
@@ -138,74 +123,19 @@ func (s *KeyCronService) runAsLeader() {
case <-ticker.C:
s.submitValidationJobs()
case <-heartbeat.C:
// Renew the lock to prevent it from expiring during long-running tasks
logrus.Debug("KeyCronService: Renewing leader lock.")
err := s.Store.Set(leaderLockKey, []byte(time.Now().String()), leaderLockTTL)
if err != nil {
logrus.Errorf("KeyCronService: Failed to renew leader lock: %v. Relinquishing leadership.", err)
return // Relinquish leadership on renewal failure
}
case <-s.stopChan:
return // Service stopping
}
}
}
// processResults consumes results from the validation pool and updates the database.
func (s *KeyCronService) processResults() {
defer s.wg.Done()
keysToUpdate := make(map[uint]models.APIKey)
ticker := time.NewTicker(30 * time.Second)
defer ticker.Stop()
for {
select {
case result, ok := <-s.Pool.ResultsChannel():
if !ok {
s.batchUpdateKeyStatus(keysToUpdate)
return
}
key := result.Job.Key
var newStatus string
var newErrorReason string
if result.Error != nil {
newStatus = models.KeyStatusInvalid
newErrorReason = result.Error.Error()
} else {
if result.IsValid {
newStatus = models.KeyStatusActive
newErrorReason = ""
} else {
newStatus = models.KeyStatusInvalid
newErrorReason = "Validation returned false without a specific error."
}
}
if key.Status != newStatus || key.ErrorReason != newErrorReason {
key.Status = newStatus
key.ErrorReason = newErrorReason
keysToUpdate[key.ID] = key
}
case <-ticker.C:
// Process batch on ticker interval
if len(keysToUpdate) > 0 {
s.batchUpdateKeyStatus(keysToUpdate)
keysToUpdate = make(map[uint]models.APIKey)
}
case <-s.stopChan:
// Process any remaining keys before stopping
if len(keysToUpdate) > 0 {
s.batchUpdateKeyStatus(keysToUpdate)
}
return
}
}
}
// submitValidationJobs finds groups and keys that need validation and submits them to the pool.
func (s *KeyCronService) submitValidationJobs() {
logrus.Info("KeyCronService: Starting validation submission cycle.")
@@ -272,32 +202,3 @@ func (s *KeyCronService) updateGroupTimestamps(groups map[uint]*models.Group, va
logrus.Errorf("KeyCronService: Failed to batch update last_validated_at for groups: %v", err)
}
}
func (s *KeyCronService) batchUpdateKeyStatus(keysToUpdate map[uint]models.APIKey) {
if len(keysToUpdate) == 0 {
return
}
logrus.Infof("KeyCronService: Batch updating status for %d keys.", len(keysToUpdate))
var keys []models.APIKey
for _, key := range keysToUpdate {
keys = append(keys, key)
}
err := s.DB.Transaction(func(tx *gorm.DB) error {
for _, key := range keys {
updates := map[string]any{
"status": key.Status,
"error_reason": key.ErrorReason,
}
if err := tx.Model(&models.APIKey{}).Where("id = ?", key.ID).Updates(updates).Error; err != nil {
logrus.Errorf("KeyCronService: Failed to update key ID %d: %v", key.ID, err)
}
}
return nil
})
if err != nil {
logrus.Errorf("KeyCronService: Transaction failed during batch update of key statuses: %v", err)
}
}