It would seem good to have HA for the master nodes so we can reboot masters without taking down the cluster. In that case, we should run the agent on the master nodes to keep the cluster economical, and take full advantage of all the extra resources allocated for master nodes.
It would seem good to have HA for the master nodes so we can reboot masters without taking down the cluster. In that case, we should run the agent on the master nodes to keep the cluster economical, and take full advantage of all the extra resources allocated for master nodes.
Here's the structure I have in mind: