Repository navigation
Update the backup-and-recovery guide for crash-safe restores and the 5.4 peer refusal - #794
Conversation
… new restore_backup options for 5.4 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…ers, stale limitations and 5.4 release notes Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
There was a problem hiding this comment.
Code Review
This pull request updates the documentation for database backup and recovery operations to align with the v5.4.0 release. The changes cover online restores using target_database, replication peer restrictions, pre-copy verification checks, crash survival behaviors, and new parameters like accept_incompatible_backup. The review feedback recommends formatting adjustments for the <VersionBadge> component, advising against its mid-sentence usage in prose and correcting its type attribute for new features.
🚀 Preview DeploymentYour preview deployment is ready! 🔗 Preview URL: https://preview.harper-documentation.harperfabric.com/pr-794 This preview will update automatically when you push new commits. |
|
Heads-up: the v5 reference and the 5.4 release notes now live in the code repos (documentation#788, harper#3944, harper-pro#1237).
— Claude Opus 5.5 |
…rper's docs/ The reference and release-note changes this branch carried moved to HarperFast/harper#3920, which carries its own docs since the v5 reference moved into harper. Only the Learn page change stays here. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
🚀 Preview DeploymentYour preview deployment is ready! 🔗 Preview URL: https://preview.harper-documentation.harperfabric.com/pr-794 This preview will update automatically when you push new commits. |
🧹 Preview CleanupThe preview deployment for this PR has been removed. |
⊙ Problem
Publish every restore_backup through one crash-safe protocol (harper#3920) changes what
restore_backupdoes in 5.4. The backup-and-recovery guide still saystarget_databaseneeds the server stopped and that an interrupted restore always leaves the database unloadable. Its replicated-cluster runbook would also now fail, because a restore refuses to run while the node lists peers.The v5 reference and the 5.4 release notes have moved into harper's
docs/, so this PR's reference and release-note changes now ship in harper#3920 itself. Only the Learn guide, which stays in this repo, is left here.❓ Your call: is a runbook step that removes the node's peers the right guidance?
harper#3920 refuses a restore while
system.hdb_nodesorreplication.routesnames a peer, even one the network has isolated. The runbook now says to remove the node's peers before restoring and add them back at re-entry. The alternative is to document an override, which the code does not have.💡 Solution
target_databaseworks online as of v5.4.0, what an interrupted restore leaves (as it was, restored, or blocked), and that a restore refuses on a node with replication peers.✅ Verification
mainonly in this page.🤖 Generated with Claude Code
Related PRs: #788 overlaps (moved the reference and 5.4 notes this PR used to edit into harper; those edits now ship in harper#3920)
Complexity: easy
Review-Coverage: authored=claude; ran=gemini,codex; adjudicated=domain; blocked=cursor-composer(auth); declined=cursor-grok,muse; rounds=2; full=1 @ 8e3b081
Review-Attention: skim ~2m (decisions: peer-refusal-guidance, restore-option-scope) @ 8e3b081