fix(remote): multi-machine spawn + dated target names - #40
Open
cgpadwick wants to merge 3 commits into
Open
Conversation
…ot content On spark, `saage remote spawn` launched fine but ssh never came up and the box was terminated after the 300s wait. Root cause: ensure_ssh_key treated an existing Lambda key NAMED 'saage-remote' as ours, but that name was registered from a different machine — the instance booted authorized for the other machine's key, so every ssh attempt was Permission denied (silently, under BatchMode) until wait_ssh gave up. - ensure_ssh_key now matches key CONTENT (type+blob, comment ignored) and returns the usable name; on a name collision with a different key it registers a stable 'saage-remote-<fp8>' variant instead of shadowing the other machine's entry. launch() uses the returned name. - wait_ssh fails fast on 'Permission denied' with a pointed message — auth rejection is deterministic, looping 300s only buried the cause. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
lambda-<hhmm> collided across days and told you nothing about the box's age — half the reason stale targets were unidentifiable. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
Remote-spawn reliability fixes plus a provider bump.
ensure_ssh_keynow matches Lambda keys by CONTENT (type+blob) rather than by name. A key registered under the namesaage-remotefrom a different machine no longer shadows ours; on a name collision with different content it registers a stablesaage-remote-<fp8>variant.wait_sshfails fast onPermission deniedinstead of burning the full 300s wait on a deterministic auth rejection.lambda-<yyyymmdd-hhmmss>so targets are unique across days and carry the box's age (stale-target cleanup was impossible with the oldlambda-<hhmm>scheme).Test plan
pytest -q(offline suite)saage remote spawnsucceeds from a second machine (spark) where it previously terminated the box after the ssh wait timed out.🤖 Generated with Claude Code