You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Jan 30, 2020. It is now read-only.
This feature has been requested several times in one form or another. The general use case is "run one copy of a particular unit on every single machine in the cluster".
In implementation terms, the basic idea is that there is a static manifest of unit in fleet's registry (etcd), and every fleet agent watches this central depository, cloning and starting units locally. Hence, it completely bypasses the fleet scheduling engine.
Various supporting changes will be necessary before this can be implemented.
The text was updated successfully, but these errors were encountered:
This feature has been requested several times in one form or another. The general use case is "run one copy of a particular unit on every single machine in the cluster".
In implementation terms, the basic idea is that there is a static manifest of unit in fleet's registry (etcd), and every fleet agent watches this central depository, cloning and starting units locally. Hence, it completely bypasses the fleet scheduling engine.
Various supporting changes will be necessary before this can be implemented.
The text was updated successfully, but these errors were encountered: