Our open source version already supports some forms of scaling/clustering today: e.g., read-only replicas for hot standbys and increasing query throughput. Timescale also supports hundreds of thousands of row inserts (millions of metrics) per second on a single node, and allows users to elastically add disks to scale up single-node capacity to ~100TB.

For scaling beyond those insert and storage limits, we're currently working on full, scale-out clustering. We'll have more to say about that soon.

I _really_ love that you all have built something that works really well when scaled vertically, and not kneecapped the whole thing in search of horizontal scale chasing the 1%-ers.