Rewrite Flycast blueprint to drop GPU/Ollama example#2394
Merged
Conversation
GPU Machines are deprecated as of 2026-08-01, so the Ollama walkthrough needed to go. Reframed the guide around a generic private HTTP service using the public nginxdemos/hello image and the modern fly launch --flycast flag, so no external example repo is needed. Also swapped the Fly Postgres mention for Managed Postgres and refreshed related reading to drop the GPU/Ollama blog posts.
The link pointed at /docs/reference/services/#ip-addresses, which no longer exists. The relevant content lives at /docs/networking/services/#anycast-ip-addresses.
The image listens on port 80, but fly launch defaults the internal port to 8080, which leaves the app unreachable over Flycast. Pass --internal-port 80 explicitly and explain why.
fly launch enables force_https in [http_service] by default, which makes the proxy redirect plain HTTP to HTTPS. Flycast addresses don't have public TLS certs, so the redirect breaks the curl example. Add --no-deploy to the launch step, instruct the reader to flip force_https to false, then deploy.
The default 256MB shell Machine gets OOM-killed midway through 'apt install curl iputils-ping dnsutils'. 1GB has plenty of headroom and avoids a confusing 'Killed' message that makes readers think the walkthrough is broken.
Roadmaster
approved these changes
May 19, 2026
Contributor
Roadmaster
left a comment
Roadmaster
left a comment
There was a problem hiding this comment.
Awesome! Thanks! One small nitpick.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
GPU Machines are being deprecated on 2026-08-01, and the Flycast blueprint at
blueprints/private-applications-flycast.html.mdwas built almost entirely around an Ollama-on-GPU walkthrough. The Flycast concepts themselves are GPU-independent, so this rewrites the guide rather than removing it.nginxdemos/helloimage, so no external example repo is neededfly launch --no-deploy+fly ips allocate-v6 --private+fly deployflow for the modernfly launch --image ... --flycast/docs/reference/services/#ip-addresses→/docs/networking/services/#anycast-ip-addresses); the old path no longer resolveswalkthroughtostyles/config/vocabularies/fly-terms/accept.txtso Vale stops flagging it as a typo (it's already used elsewhere in the docs)Walkthrough gotchas baked into the recipe
--internal-port 80onfly launch:nginxdemos/hellolistens on 80, but Fly Launch defaults to 8080force_https = falseinfly.toml: Flycast addresses don't have public TLS certs, so the default 301-to-HTTPS breaks the demo's plain HTTP requests--vm-memory 1024on the shell Machine: the 256MB default OOM-kills duringapt install -y curl iputils-ping dnsutils