Correct what the telemetry guide says about egress and credentials - #504
Open
dennis-upbound wants to merge 1 commit into
Open
dennis-upbound wants to merge 1 commit into
dennis-upbound wants to merge 1 commit into
Conversation
Three claims in one paragraph of the shipped guide are wrong, and the third is about where a user's credential is held. "Your clusters reach the control plane, and only the control plane reaches your backend." Modelplane composes a collector onto each inference cluster and no collector onto the control plane. Each cluster exports to the sinks itself. "A cluster with no route to your observability stack still reports." It does not. Without a route to the backend it exports nothing. "The backend's credential lives in one place instead of on every GPU cluster." The opposite is true, and deliberately so: compose-serving-stack resolves the Secret a sink names and composes a copy onto every cluster that runs a collector, because the collector mounts it there. A reader deciding whether to give Modelplane a vendor token would have read this and been wrong about where the token ends up. The guide elsewhere describes the copy correctly, so this paragraph is the one place that contradicts both the code and the rest of the page. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Signed-off-by: Dennis Ramdass <dennis@upbound.io>
|
Docs preview: https://modelplane-docs-pr-504.vercel.app (ready once the site's Content workflow finishes) |
3 of 4 tasks
tr0njavolta
approved these changes
Oct 5, 2026
This branch has not been deployed
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Description of your changes
One paragraph of the shipped v0.5 telemetry guide makes three claims, and all three are false. Found while researching #476.
compose_collector(fn.py#L772) is called only from the ServingStack path and composes onto each inference cluster. Each cluster's collector exports to the sinks directly.The third is the one worth fixing quickly: someone deciding whether to hand Modelplane a vendor token reads that sentence and gets the wrong answer about where the token ends up.
The rest of the page already describes the copy correctly (
docs/content/platform/telemetry.md:101), so this paragraph contradicts both the code and its own guide. I also dropped a colon on that line thatai-tells.ColonUsageflags locally.Wants a backport to
release-0.5— the claim is in the published 0.5 docs.I have:
nix flake check(or./nix.sh flake check) and made sure it passes.Added or updated tests covering any composition function changes.Docs only.git commit -s.🤖 Generated with Claude Code