Dkron is a distributed and fault-tolerant task scheduling system, written in Go. It is an excellent replacement for traditional cron when reliability is needed : where an isolated cron depends on a single machine, Dkron distributes scheduling across a cluster with no single point of failure.

High availability relies on the Raft protocol for leader election and the Gossip protocol (Serf) for node communication : if the leader fails, another node automatically takes over, without human intervention. Jobs are scheduled using a cron-like syntax and executed via shell or HTTP executors, with chaining and node targeting by tags.

On the operational side, Dkron provides a simple web interface and JSON API : execution history, statuses, and notifications (email, webhooks, Slack…) in case of success or failure. This makes it an ideal automation foundation for critical scheduled processing. Hosted and monitored by DINAO, it offers this robustness without the administrative burden of managing a homegrown cluster.