You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Local repository creation via the admin API failed with SSH errors (#356). POST /api/v1/clients/{id}/repositories ran borg init over the repo's SSH path using the agent's SSH identity and a hardcoded port 22, so it failed on any install where the SSH port isn't 22 (including the docker-compose default of 2222). Repo init now runs directly on the local path via the root helper — same flow the web UI already used — and repo ownership is set so agent backups can acquire the repo lock. Thanks @Rhonstin for the detailed report.
Scheduler: a hung run can no longer exhaust the database connection pool. Scheduler runs are now guarded by a lock file so only one instance runs at a time.