During this time, I will be restarting Fire to deploy bug fixes while also doubling the cluster count to 4 and having a shard per cluster.
This will hopefully reduce delays during reboots and will allow me to have finer control over which servers get an update first.
Fire will be offline for a couple minutes while I make these changes. I will eventually have a way to change the cluster & shard counts without complete downtime but did not find it was necessary as of yet.
