Scale out and scale in automatically based on a metric such as CPU utilization rate. Scale out should add a specified number of instances, run build and deploy, and then automatically add it to the cluster. Scale in should remove the instance from the cluster and destroy it.
A public API for automating operations like: * Creation/deletion of Render services * Addition/removal of custom domains * Creation/deletion of databases for users
Support for serverless scripts
I would like to be able to replace Amazon Lambda with a serverless setup hosted on Render.
Disks for Cron Jobs
Certain cron jobs need to persist data on disk between runs, and Render Disks work great for this.
Respect .nvmrc in addition to .node-version
I use nvm, Node Version Manager, to manage my Node.js version when developing. It'd be nice to have Render respect the version listed in the .nvmrc file that nvm uses to determine the node version when deploying instead of having to use a separate .node-version file.
Allow disks on services with multiple instances
We've been moving our main datastore off of Couchbase for a few years now, but I've still got to run it. Right now I have it running on four or five very large instances on EC2. Each instance has its own 1TB or so of disk space, and the database runs in cluster mode, with each of the nodes doing service discover to find each other. It'd be very nice to run this on Render. I could make a distinct service per node and bootstrap the cluster by hand, but it might be nice to be able to just configure the number of instances, and have each instance get its own disk space automatically.
Download data from Render disks
Right now, I have to upload disk data to a non-Render location if I want to inspect it locally. It would be nice to have an easy way to download the data stored on the disk to a local machine.
Ability to install Python libraries on an Elixir service
We're using an Elixir service to host our Phoenix app. We need to make use of some Python libraries, ideally we want to be able to run Python scripts from within Elixir code using Thrift or Erlports. I've tried adding pip3 commands to build.sh , but the libraries aren't persisted to the service. They are installed to a directory called /opt/render/.local/lib/python3.7/site-packages which isn't present on the server after it's live. I know we can use Docker, we just want to use it if possible.
I would like to be able to have multiple deploys running at once. Let's say deploys take 10 minutes, and I deploy Version A at 6:15pm, and Version B at 6:20. With the current behaviour, it would deploy Version B at 6:30, canceling the deploy of A. I would like to be able to have Version A running from 6:25-6:30, then Version B from 6:30 onwards.