Snapshot Browser won’t expand after deleting and re-adding the same repository under a new name
Describe the bug My understanding is that “repository name” is only a Backrest label, restic itself lets you connect to a repo by URL/credentials, not by a human-readable name. After I delete a repository in Backrest and re-add the same underlying restic repo with a different Backrest repo name, the connection test succeeds (“Connected successfully to rest:http://x.x.x.x:8000/user and found an existing repo”), and maintenance actions (index, unlock, prune, check) work. However, the Snapshot Browser will not expand/load for that repository. No reverse proxy is involved; I connect directly over VPN.
To Reproduce Steps to reproduce the behavior:
- Create a repository in Backrest, name it “RepoA”.
- Create a snapshot
- Delete “RepoA” in Backrest.
- Re-add the same underlying repository but with a different display name, e.g., “RepoB”.
- Click Test configuration → success message appears.
- Open the repository and try to expand Snapshot Browser.
Expected behavior Snapshot Browser expands and lists files backed up.
Actual behavior Snapshot Browser does not expand/load. Other actions (index, unlock, prune, check) succeed.
Screenshots
Platform Info
- OS and Architecture: Windows 10 x64
- Backrest Version: 1.9.2
- Repository in REST server
Additional context This looks like a UI/state issue tied to the Backrest display name or cached repo metadata. The backend operations succeed, suggesting the underlying repo is reachable and valid.
Please let me know if additional information is needed to address this issue, or if it's not a bug and I'm doing something wrong 😅
Thanks for the report -- this looks like a real issue. Appreciate your including a screenshot of the request contents, looks like it's because the API works with repo ID, but the storage model is based on the repo GUID. So you're seeing backup operations indexed before the repo was renamed.
When you try to browse them, they call the API with the old repo name. A database reset would fix the issue once snapshots are reindexed, but it's something I should fix.
The right way to address this is for me to update the API to refer to repos by their GUIDs.
Saw this same behavior after I migrated a Windows 10 laptop to openSUSE Leap 16. I installed Backrest v1.10.1 on the new Linux install, and added a repo with the path to the Win 10 repo on my USB backup drive.
No re-indexing, pruning or checking worked to allow me to browse the snapshots, so I created a Linux repo, performed a tiny backup to confirm operation, opened a terminal session, made sure restic was installed, and copied the snapshots to the Linux repo using restic commands. After re-indexing, pruning and checking, browsing snapshots now works.
Here is my procedure:
Computer name = Sunny User name = ferocious New Linux repo = LinuxRepo Old Windows repo = WindowsRepo Mount path for USB drive = /run/media/ferocious/SysBackup
Command contents = ferocious@Sunny:~> restic -r /run/media/ferocious/SysBackup/LinuxRepo copy --from-repo /run/media/ferocious/SysBackup/WindowsRepo
The restic command asks for passwords for both the target and source repos before beginning the copy.