Currently glb-healthcheck runs health checks on all backends in the forwarding table, even when they are marked inactive and it doesn't make sense to do so.
The forwarding table documentation states:
* `inactive` - the backend is completely ignored
This PR disables the healthchecking of inactive backends to act the same as the director itself. Health checking these backends will also cause needless reloads of the forwarding table, since inactive ones are filtered out downstream in glb-director-cli as well.
Currently glb-healthcheck runs health checks on all backends in the forwarding table, even when they are marked
inactiveand it doesn't make sense to do so.The forwarding table documentation states:
This PR disables the healthchecking of
inactivebackends to act the same as the director itself. Health checking these backends will also cause needless reloads of the forwarding table, sinceinactiveones are filtered out downstream in glb-director-cli as well.