Update the config

Deploying a change is the same loop you already know — edit, test, publish — followed by assigning the new version to your fleet. Collectors pick it up on their next phone-home and converge.

Roll out a new version

1. Edit the configuration in the editor and test it.
2. Publish the new version.
3. Assign that published version to the fleet as its configuration.
4. Watch the fleet on the dashboard: instances move from Pending pickup → Applying → In sync as they receive and apply the new config.

Because collectors pull from Telflo, you don't touch the hosts — the rollout happens over OpAMP.

Roll back

To revert, assign an earlier published version to the fleet. The same convergence applies in reverse: instances pick up the older version and report In sync once they've applied it. Keeping clean version history (with notes) makes choosing the right rollback target straightforward — see Versions & publishing.

If an apply fails

If an instance reports Apply failed, the collector rejected the configuration:

1. Open the configuration and re-check it — validate and test it against sample data.
2. Check the collector's own logs on the host for the rejection reason.
3. Fix the issue, publish a corrected version, and assign it to the fleet.

Next

• Spot instances that haven't caught up yet — Configuration drift.
• Keep your install token healthy — Tokens & security.