You guys really need to do this on GCP where there's zero Postgres competition.
Since GCP will be overtaking AWS in like 2-3 years (hah), good to get a head start now. ;)
Since GCP will be overtaking AWS in like 2-3 years (hah), good to get a head start now. ;)
You may need something that's a little more cost accessible before it's widely useful ;)
Or you may want to allow a one-click solution that I can deploy on my resources now and later seamlessly move to your hosted solution, when I need to scale.
That kind of growth would make the growth of AWS look like a local corner-store.
Re Citus:
Considering that you get 2 GB of bandwidth per core on GCE, that makes a lot of sense for a distributed DB like Citus where there's a lot of inter-node traffic.
That other point there is that the maximum amount of RAM for a VM is 208Gb, which is not enough to scale up a Postgres instance. So a barrier to entry is porting your entire data layer to Datastore -- that's a significant engineering challenge that most people are unlikely to take up.
So Citus would fit in really well. IMHO focusing on AWS (where Amazon are providing their own DB services) is missing an opportunity on GCP (where Cloud SQL is MySQL based and fairly limited).