Skip to content

GH-572: Document Config Server behavior when Git backend is unavailable#3251

Open
won-seoop wants to merge 1 commit into
spring-cloud:mainfrom
won-seoop:gh-572-document-backend-unavailability
Open

GH-572: Document Config Server behavior when Git backend is unavailable#3251
won-seoop wants to merge 1 commit into
spring-cloud:mainfrom
won-seoop:gh-572-document-backend-unavailability

Conversation

@won-seoop

Copy link
Copy Markdown

Summary

Adds a new "Behavior When the Git Backend is Unavailable" section to git-backend.adoc covering the four scenarios requested in the original issue:

  1. At Startup — explains the difference between lazy-clone (default, silent failure) and clone-on-start: true (fail-fast), and points to the health indicator for post-startup detection.
  2. After a Successful Startup — explains that the Config Server falls back to its local working copy and continues serving clients while the remote is down; fetch attempts fail silently and the local copy is used until the remote recovers.
  3. Local Copy Drift and Force Pull — explains the risk of local/remote divergence and cross-references the existing force-pull section.
  4. Effect on Client Refresh — explains what happens when a client refresh is triggered while the backend is unreachable (local copy fallback, or error if no local copy exists).

Related issue

Closes #572

🤖 Generated with Claude Code

…is unavailable

Add a new section "Behavior When the Git Backend is Unavailable" with
four sub-sections explaining the different failure scenarios:

- Startup: lazy-clone vs clone-on-start fail-fast behavior
- Runtime: local working copy fallback when remote is down
- Local copy drift: how force-pull prevents stale local state
- Client refresh: what happens when a refresh is triggered while the
  backend is unreachable

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Add docs for Config Server behavior when backend is not accessible

2 participants