Durable job queue
Publishing, email and automation jobs run through pg-boss, a PostgreSQL-backed job queue. Failed jobs are retried automatically with exponential backoff.
Pyron uses a durable, queue-backed architecture to ensure that publishing, email delivery and automation work reliably — even when individual components fail temporarily.
Publishing, email and automation jobs run through pg-boss, a PostgreSQL-backed job queue. Failed jobs are retried automatically with exponential backoff.
If the job queue is temporarily unavailable when you publish, your deployment is saved to a durable outbox. The worker recovers and processes it automatically.
Social publishing and external API calls use idempotency keys to prevent duplicate actions during retries.
Every successful website release can be restored in one click. We store full deployment snapshots, not just diffs.
Production deployments include structured logging with request IDs and error tracking. Operators can diagnose failures without exposing stack traces to users.