Search
13 results for “retry”
Docs8
- DocsCore conceptsRetries and backoff
- DocsPlatformRate limitsRetry-After header429 responses include Retry-After in seconds; the SDKs wait and retry automatically.
- DocsCore conceptsRetries and backoffCustom retry policiesSet maxAttempts and the backoff curve once per task instead of on every job.
- DocsCore conceptsRetries and backoffNonRetryableErrorThrow it to fail a job immediately when a retry can't help, like a deleted user.
- DocsPlatformWebhooksWebhook delivery retriesFailed webhook deliveries retry for up to 3 days with exponential backoff.
- DocsCore conceptsRetries and backoffExponential backoffDelays, jitter and the maximum attempt count.
- DocsCore conceptsRetries and backoffMaximum attemptsJobs retry up to 5 times by default. Raise or lower the limit per job with maxAttempts.
- DocsCore conceptsRetries and backoffDead-letter listJobs that run out of attempts land here; inspect the last error and retry them in bulk.
API reference2
Guides3
- GuidesMake jobs idempotentWhen retries should stopA declined card won't succeed on attempt four. Retry only failures that clear on their own.
- GuidesMigrate from cron jobsFailures retryUnlike cron, a failed Halyard job retries with backoff instead of waiting for its next slot.
- GuidesMake jobs idempotentWhy jobs run twiceAt-least-once delivery means a retry can repeat a side effect. Design every job to tolerate it.