Hey Ollie,
Was just getting back to your email!
In short, every configuration change on a cluster without High Availability enabled leads to downtime on the node, as it requires reindexing all data and sometimes even infrastructure changes.
Since your setup is a single-node cluster, this naturally makes the entire cluster unavailable during the change. If high availability is critical for you, it would be import to look into the High Availability feature in Typesense Cloud. It provisions a 3-node cluster, allowing configuration changes to be applied one node at a time, without causing downtime.
As for how long the reindexing takes: that depends on the size of your dataset and your node’s configuration. Since it reindexes everything, larger datasets or lower capable nodes will take longer to complete the process for each node.