Can your team operate Forgejo in production?

Check the work around the forge before deciding to run it yourself. The install is only the first line item.

Forge operations readiness checklist

Backups and restores

Backups run automatically on a documented schedule. The job covers repositories, database state, attachments, LFS, configuration, and secrets. Backups are encrypted, off-site, and retained long enough for your risk model. A local disk snapshot alone is not a disaster recovery plan. Someone has restored the forge from backup in the last quarter. A backup is only real when the restore path has been tested.

Updates and security

A named owner tracks Forgejo releases and security advisories. The owner knows when to update, when to wait, and how to roll back. There is a maintenance window and rollback plan for upgrades. Production source control should not be updated by improvisation. The host, database, reverse proxy, and runner machines are patched. Forgejo is one part of the operating surface, not the whole surface.

Monitoring and incidents

HTTP, SSH, database health, disk, memory, and CPU are monitored. Disk growth and queue failures are common ways small forges become urgent. Alerts route to someone who can act outside office hours. If nobody is responsible, the alert is just noise. There is a short runbook for outage, restore, and customer communication. Even a small team needs to know who makes the call during a forge incident.

CI, access, and migration

Runner capacity is patched, isolated, and sized for real workloads. CI turns a forge into compute infrastructure with a different risk profile. User access, deploy keys, webhooks, and bot accounts are reviewed. A forge becomes sensitive because it controls source, secrets paths, and deploy flow. The team knows how to import, export, and leave without losing history. Open exit matters before the first outage or procurement change.

Readiness score

0/12

Fjord is likely the lower-maintenance path.

You can still operate Forgejo yourself, but the missing pieces are production responsibilities, not setup chores.

Start with managed Forgejo Read the migration guide

How to read the score

A low score does not mean Forgejo is hard to run. It means source control is production infrastructure. If your team wants the control but not the operating job, Fjord runs the dedicated Forgejo instance, runners, updates, TLS, monitoring, and Team/Pro backups for you.

Choose the operating model before the install.

Run Forgejo yourself if your team wants the responsibility. Use Fjord if you want a dedicated upstream Forgejo operated for you.

See managed Forgejo hosting Compare self-hosting →See pricing →Create an account →