update readme

Author: sohel2020

pharia-ai-backup-restore/README.md

@@ -63,34 +63,48 @@ psql -h localhost -p 5432 -U pharia_user -d dev -c "SELECT version();"
 
 ### ⬆️ Perform Upgrade
 
-```bash
-helm upgrade pharia-ai ./pharia-ai-chart --namespace pharia-ai
-```
+Follow your standard deployment process to upgrade Pharia AI.
 
 ### ✅ If Upgrade Succeeds
 
 Test your application and you're done!
 
 ### ⚠️ If Upgrade Fails: Rollback
 
+> **Important:** Always stop application pods before restoring databases to prevent data corruption and ensure a clean restore.
+
 ```bash
-# 1. Restore databases
+# 1. Scale down application pods to prevent database writes during restore
+kubectl scale deployment --all --replicas=0 -n pharia-ai
+
+# 2. Verify all pods are terminated
+kubectl get pods -n pharia-ai
+# Wait until no application pods are running (only database pods should remain)
+
+# 3. Restore databases
 ./bin/pharia-backup.sh db restore all
 
-# 2. Restore secrets
+# 4. Restore secrets
 ./bin/pharia-backup.sh secrets restore --latest -f -n pharia-ai
 
-# 3. Rollback Helm deployment
-helm rollback pharia-ai -n pharia-ai
+# 5. Scale deployments back up (adjust replica counts as needed for your setup)
+kubectl scale deployment --all --replicas=1 -n pharia-ai
 
-# 4. Verify pods are running
-kubectl get pods -n pharia-ai
+# 6. Verify pods are running and healthy
+kubectl get pods -n pharia-ai -w
 ```
 
+> **💡 Note:** If your deployments have different replica counts, you can scale them individually:
+> ```bash
+> kubectl scale deployment <deployment-name> --replicas=<count> -n pharia-ai
+> ```
+
 ## 📝 Common Commands
 
 ### 🗄️ Database Operations
 
+> **💡 Tip:** Before restoring databases, always scale down your application pods to zero replicas. This prevents data corruption from concurrent writes during the restore process.
+
 ```bash
 # Backup all databases
 ./bin/pharia-backup.sh db backup
@@ -140,13 +154,20 @@ ls -lt database-backups/ | head -3
 ls -lt secrets-backups/ | head -3
 
 # === UPGRADE ===
-helm upgrade pharia-ai ./pharia-ai-v2.0 -n pharia-ai
+# Follow your standard deployment process
 
 # === IF ROLLBACK NEEDED ===
+# Step 1: Stop application to prevent writes during restore
+kubectl scale deployment --all --replicas=0 -n pharia-ai
+kubectl get pods -n pharia-ai  # Wait for pods to terminate
+
+# Step 2: Restore data from backups
 ./bin/pharia-backup.sh db restore all
 ./bin/pharia-backup.sh secrets restore --latest -f -n pharia-ai
-helm rollback pharia-ai -n pharia-ai
-kubectl get pods -n pharia-ai
+
+# Step 3: Bring application back online
+kubectl scale deployment --all --replicas=1 -n pharia-ai
+kubectl get pods -n pharia-ai -w  # Watch pods come back up
 ```
 
 ## 🔧 Troubleshooting