fix: Implement distribute locker in Redis for cron jobs

backend/src/features/backup/README.md

@@ -111,7 +111,23 @@ Backups are stored in `/app/data/backups/` (mapped to `./data/backups/` on host)
 
 Jobs are registered in `backend/src/core/scheduler/index.ts`:
 - Backup check: Every minute
-- Retention cleanup: Daily at 4 AM
+- Retention cleanup: Daily at 4 AM (also runs after each scheduled backup)
+
+### Distributed Locking
+
+Scheduled backups use Redis distributed locking to prevent duplicate backups when multiple backend containers are running (blue-green deployments).
+
+**Lock behavior:**
+- Lock key: `backup:schedule:{schedule_id}`
+- Lock TTL: 5 minutes (auto-release if container crashes)
+- Only one container creates the backup; others skip
+
+**Retention cleanup:**
+- Runs immediately after each successful scheduled backup
+- Deletes backups exceeding the schedule's retention count
+- Also runs globally at 4 AM daily as a safety net
+
+See `backend/src/core/scheduler/README.md` for the distributed locking pattern.
 
 ### Admin Routes