Importing SVN Repos

subforge can bring an existing SVN repository in from another server, preserving the full history — revision numbers, authors, dates, and commit messages.

Starting an import

Go to New repo, fill in the name and choose the target workspace, then select one of the import modes:

• Import from URL — one-shot copy. The new repo is fully writable after the import completes. Credentials are used only during the import and are not stored.
• Synced mirror — continuous sync. subforge polls the upstream URL on a configurable interval and pulls in new revisions automatically. The repo is permanently read-only; credentials are stored on disk so the syncer can keep polling.

The source URL must be accessible from subforge's server — HTTP, HTTPS, and svn:// are all supported.

UUID preservation

SVN repositories have a UUID — a unique identifier embedded in the repository metadata. If you enable Preserve UUID during import:

• Existing working copies of the original repo can be repointed at subforge using svn relocate, without needing a fresh checkout.
• The UUID is locked to the imported value and cannot be changed later.

Leave UUID preservation off if you want subforge to assign a fresh UUID. In that case, all existing working copies will need a fresh svn checkout after the import.

Import progress

Imports run in the background. While in progress, the repo page shows a progress indicator. Large repositories may take several minutes or longer depending on revision count and network speed.

You can leave the page and come back — the import continues regardless. Refresh the repo page to see the latest status.

After a successful import

Once complete, the repo is fully functional. If the UUID was preserved, the Relocate button appears on the Code tab with the svn relocate command. Distribute this command to your team so they can point their existing working copies at subforge.

Failed imports

If an import fails (network error, authentication failure, etc.), the repo page shows an error message. You can:

• Retry — attempt the import again with the same settings. Useful for transient network errors.
• Cancel — mark the repo as failed and stop retrying. You can then delete the repo or try again with a different URL.

Post-commit hooks during import

Post-commit hooks are suspended during import to avoid flooding integrations with historical commits. They resume for new commits once the import finishes.

Limitations

• The source server must allow anonymous read or you must provide credentials.
• Very large repos (millions of revisions) may hit timeouts. Contact support if this is a concern.