8000 cron: configurable retry policy for failed runs · Issue #115 · acidtib/jiji · GitHub
[go: up one dir, main page]

Skip to content

cron: configurable retry policy for failed runs #115

Open

Description

@acidtib

Summary

A failed cron run is never retried; it waits for the next scheduled tick. CronConfig (crates/jiji-config/src/schema.rs) has schedule, command, timezone, timeout, overlap, and missed_runs, with no retry setting, and the agent scheduler has no retry path. This is a feature addition, not a defect: the website documents "Failed runs do not retry automatically."

Scope

  • A configurable retry policy per job (e.g. attempt count and backoff).
  • Record each retry as its own run with a distinct cause (scheduled, manual, retry), visible in jiji service cron status/logs.
  • Define how retries interact with overlap: forbid and the next scheduled tick.

Done when

The cron reference on the website ("Current Limits") and crates/jiji-config/src/jiji.yml are updated in the same change.

Code: crates/jiji-config/src/schema.rs, crates/jiji-agent/src/cron.rs, crates/jiji-agent/src/cron_exec.rs, crates/jiji-agent/src/scheduler.rs.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions

      0