Recover management access

If another Admin can sign in, use user management. Otherwise, a host administrator can run recovery inside the running CDelivery main container.

The commands require existing platform data and its master key. A fresh container without those resources cannot substitute for the deployed instance, and login recovery cannot repair a lost key or application volume.

Find the container

On the server running the main program:

docker ps --format '{{.ID}} {{.Names}} {{.Image}}'

Identify the main CDelivery container and replace CONTAINER_ID below. Do not choose a gateway or collector.

Reset an existing account

docker exec -it CONTAINER_ID /entrypoint rescue admin@example.com

This prints a temporary password and disables that account’s two-factor authentication. It does not promote the user to Admin or revoke API keys. Sign in, change the password, re-enroll two-factor authentication, and review keys.

Add an administrator

For an email that does not already exist:

docker exec -it CONTAINER_ID /entrypoint add-admin admin@example.com Alex Chen

The command prints a temporary password. An existing email conflicts; rescue that account instead. Remove unneeded temporary accounts after recovery.

Issue an API key

docker exec -it CONTAINER_ID /entrypoint create-api-key admin@example.com recovery Reader

Optionally append --expires-at with a future RFC 3339 timestamp. Roles are Reader, Operator, or Admin and cannot exceed the user’s role. Save the one-time output and revoke the key after use.

For uninitialized or missing-user errors, verify the container, deployment, and email. Do not reinstall over existing data to solve an identification mistake. Output contains credentials and must not enter public logs.