Feature Requests

Big reason I wanted to use Render was the support. However it seems noone is online over weekends or outside of 9 to 5 either. Should not be difficult to have on-call support team in place considering for production work, the issues that need support tend to be quite critical and alerting that noone can wait 2 days for Monday to hit.

When custom domains are active, provide an option to do one of the following: * Disable the onrender.com subdomain * Redirect it to one of the custom domains. * Keep the site as-is but add a noindex header to all responses from the onrender.com subdomain. https://developers.google.com/search/docs/advanced/crawling/block-indexing Benefits: * Security [1] * SEO [2] ---- [1] When using a custom domain one of the benefits is that instead of having the DNS pointing to Render, you can point it to Cloudflare (for example) as a proxy. This reduces load on the origin server and the {project}. onrender.com seems like a bit of a loose-end. Could somebody DDoS this server, bypassing cloudflare's protection? (Similar to https://render.canny.io/features/p/add-option-to-redirect-the-onrendercom-subdomain-to-the-primary-custom-domain-fo ) [2] Add x-robots-tag: noindex header to deploys without assigned custom domains. This would allow testing one's full Web site content on Render without needing to alter a repo's existing "robots.txt" file yet avoid risking the duplicate-content penalty from Google and other search vendors. Similar functionality is provided by Vercel ( https://vercel.com/docs/v2/edge-network/headers?query=robot#inlinecode ).

It would be nice to have GPU instances on Render for easy ML training.

Preview environment services can share docker build cache from its parent service. https://community.render.com/t/preview-environment-docker-build-caching/3856

While it's great to have prebuilt plans, requirements vary a lot. Some service require almost no memory and a lot of CPU or the other way around. Right now, one always has to contact support for such custom plans, would be nice to have just select CPU/memory independent from each other for advanced plans.

Allow users to set account-level spending limits for variable factors.

Proactively send slack and email notifications to users when a service utilizes > 90% of its plan limits ( CPU, RAM, Disk)

Support for selecting Postgres version 17

Enable integration with Sentry for monitoring

It would be super useful to be able to do something like this in the render.yaml file: environments: - name: my-project-prod previewsEnabled: false services: # ... envVarGroups: # ... databases: # ... - name: my-project-staging previewsEnabled: true services: # ... envVarGroups: # ... databases: # ... It's basically "just" nesting the current content of the render.yaml file into the environments array. This would enable workflow such as having prod without preview environments and then having staging with preview environments. Workflow like that is currently impossible to achieve when I want to use infrastructure as a code.