realvco Docs

Host Operations section

The third section in the HOST tab’s left column. 3 red buttons for 3 critical actions affecting the entire host.

Red = high risk = always shows a confirmation dialog.

Where to find

HOST tab → bottom-left → Host Operations section

HOST tab Host Operations section HOST tab Host Operations section
Host Operations section: 3 red buttons — Create Restore Point / Revert to Restore Point / Restart Server

The 3 buttons at a glance

ButtonFunctionWhen to use
Create Restore PointSnapshot the entire machine, can rollback laterBefore big changes / before upgrades / experimenting
Revert to Restore PointRoll back the entire machine to last snapshotFailed upgrades / settings broke something / accidentally deleted important files
Restart ServerReboot the entire VPSSystem anomalies / memory leaks / kernel-level setting changes

Create Restore Point

Host restore point = a snapshot of the entire machine at a moment. Includes OS state, all files, current settings and conversations of all AI agents.

How to do it

  1. Click Create Restore Point (red button)

  2. Confirmation dialog — says “will overwrite previous restore point” + must check confirm box

    Create Restore Point confirmation dialog Create Restore Point confirmation dialog
    Must check the agreement box before clicking Confirm

  3. Check + click Confirm

  4. Wait 2-5 minutes (snapshot in background)

What you’ll see

  • After: previous restore point time updates to now
  • Failure: dialog shows error reason

Only one restore point is kept. Each new one overwrites the old. If you want older versions:

  1. Use Export Data to download conversations and settings locally first
  2. Then “Create Restore Point”
  3. Both backup mechanisms are complementary

vs Automatic Backup

PropertyHost Restore PointAutomatic Backup (AI agent settings)
ScopeEntire machine (OS + all AIs + all data)Single AI agent’s conversations & settings
TriggerManual button clickDaily 05
auto
Retention1 (latest only)7-day rolling
DetailThis pageBackup

Revert to Restore Point

Revert = rewind the entire machine to a past moment.

⚠️ Irreversible: everything done after that moment (conversations, setting changes, files) is permanently lost.

Pre-flight checks

  1. Notify active users — conversations break, messages lost
  2. Backup what you want to keep — anything since the restore point will vanish
  3. Off-peak hours — 2-5 minutes downtime impacts work hours
  4. Confirm restore point time — the time at the top of dialog is what you’ll go back to

How to do it

  1. Click Revert to Restore Point (red button)

  2. Confirmation dialog:

    Revert to Restore Point confirmation dialog Revert to Restore Point confirmation dialog
    Red warning 'permanently lost' + checkbox required to confirm

  3. Dialog content:

    ⚠️ Revert to Restore Point 2026-04-30 07:05

    Host will revert to the 2026-04-30 07:05 restore point. Everything after that moment will be permanently lost and unrecoverable!

    Admin Panel will be offline for ~2-5 minutes during restore. Notify active users first.

    🔴 This action is risky. Tick the box to confirm.

  4. Check + click Confirm

  5. Wait 2-5 minutes (admin-panel goes offline)

  6. Refresh → back to normal screen = rollback successful

When there’s no restore point

If you never Create Restore Point’d before → Revert to Restore Point has nothing to roll back to. Dialog will say “no restore point available” (or button is disabled).

Build the habit: before any major change, click Create Restore Point first.

Why so many safeguards?

Reverting affects the entire machine, unlike “delete a file” which has a Trash. 4 layers of safeguards:

  1. Red button (visual warning)
  2. Confirmation dialog (forces info review)
  3. Forced “what you’ll lose” preview (risk reminder)
  4. Required checkbox (prevents misclick)

Restart Server

Restart Server = reboot the entire VPS. All AI agents stop, all running conversations break, all Telegram / WhatsApp bots pause. About 1-3 minutes to recover.

Pre-flight checks

  1. Notify users — conversations break
  2. Confirm no critical tasks running — check each AI agent’s Activity sub-tab for “Running” scheduled tasks
  3. Off-peak hours
  4. Note the problem you’re solving — for post-restart verification

How to do it

  1. Click Restart Server (red button)

  2. Confirmation dialog:

    Restart Server confirmation dialog Restart Server confirmation dialog
    Must check agreement box before clicking Confirm to restart

  3. Dialog says: takes 1-3 minutes / all services interrupt / checkbox required

  4. Check + click Confirm

  5. Machine reboots — admin-panel becomes unreachable

Post-restart verification checklist

Re-open admin-panel and verify:

  • HOST tab’s SYSTEM UPTIME resets to a small number (e.g. 1 min) → confirms restart
  • AI Agent List all 3 agents show green light
  • Each AI agent’s Home sub-tab shows “Running”
  • OpenClaw UI / Hermes UI respond to chat
  • Telegram / WhatsApp bots (if configured) send/receive normally
  • The original problem you were solving: solved?

What restart can’t fix

ProblemWhat to do instead
AI agent misbehaving after upgradeRestore from local backup (within 5 days)
System broken after settings changeRevert to Restore Point (if you created one)
Storage fullFree Up Space
One AI agent unresponsive but others fineThat agent’s Home sub-tab → Pause → wait 5s → Resume

How often to restart

Ideal: almost never. Linux machines running 30+ days without restart is normal.

But if you see:

  • Memory > 90% sustained
  • Reaction time obviously slow
  • Memory leak in some service

…then a restart might help.