Skip to main content

🔄 Vendo System Restart

Remotely restart multiple JuanFi vending machines from your MikroTik router using API automation. Trigger system reboots on a schedule or manually for maintenance, software updates, or troubleshooting without visiting each physical location. Coordinates with multiple vendos simultaneously and handles offline devices gracefully with error logging.

info

What this does:

  • Targets multiple vendo IP addresses (10.1.1.2, 10.1.2.2, etc.)
  • Sends HTTP POST restart command via API
  • Authenticates with API key header (X-TOKEN)
  • Logs success/failure for audit trail
  • Tolerates offline vendos without script failure

Prerequisites​

  • ✅ MikroTik RouterOS device with HTTP client capability
  • ✅ One or more JuanFi vendo devices on network
  • ✅ Vendo devices with static IP addresses
  • ✅ API credentials (vendo IPs + API key)
  • ✅ Network connectivity between MikroTik and vendos
  • ✅ Access to RouterOS console (SSH, WebFig, or WinBox)
  • ✅ RouterOS v6.48+
warning

Service interruption: Restarting vendos will interrupt customer transactions. Schedule reboots during low-traffic windows.

Configuration Steps​

Option A: Terminal Configuration​

  1. Access the terminal:

    ssh admin@your-router-ip

  2. Create restart script:

    /system script add name="vendo-restart-system" source={
        :local vendos {"10.1.1.2"; "10.1.2.2"; "10.1.3.2";};
        :local apiKey "pqa92jwxrk";
        
        :foreach vendo in=$vendos do={ 
            /do {
                :log warning "============<[[ Restarting $vendo ]]>=============";
                /tool fetch http-method=post http-header-field="X-TOKEN: $apiKey" \
                    url="http://$vendo/admin/api/restartSystem"
            } on-error={
                :log error "============<[[ ERROR! $vendo OFFLINE... ]]>=============";
            }
        }
    }
    tip
    • Update vendo IPs in the array
    • Replace API key with your actual credentials

  3. Schedule weekly maintenance restart (Sunday 2 AM):

    /system scheduler add name="vendo-restart-weekly" on-event="vendo-restart-system" \
        start-time=02:00:00 interval=7d comment="Weekly vendo restart"

  4. Create alternative script for immediate restart:

    /system script add name="vendo-restart-now" source={
        :local vendos {"10.1.1.2"; "10.1.2.2"; "10.1.3.2";};
        :local apiKey "pqa92jwxrk";
        :local timestamp [/system clock get time];
        
        :log warning "IMMEDIATE RESTART TRIGGERED at $timestamp";
        
        :foreach vendo in=$vendos do={ 
            /do {
                :log warning "Restarting $vendo";
                /tool fetch http-method=post http-header-field="X-TOKEN: $apiKey" \
                    url="http://$vendo/admin/api/restartSystem"
            } on-error={
                :log error "FAILED: $vendo";
            }
        }
    }

  5. Verify scripts:

    /system script print

Option B: WebFig Configuration​

  1. Navigate to System > Scripts:

    • Click +
    • Name: vendo-restart-system
    • Paste script from Option A step 2
    • Click OK

  2. Navigate to System > Scheduler:

    • Click +
    • Name: vendo-restart-weekly
    • On Event: vendo-restart-system
    • Start Time: 02:00:00
    • Interval: 7d
    • Click OK

Verification​

  1. Test script manually:

    /system script run vendo-restart-system

  2. Check logs:

    /log print where topics~"system"

    Should show restart attempts.

  3. Verify vendo responded:

    • Monitor physical vendo display
    • Should show reboot sequence/splash screen
    • Service down for 30-60 seconds typically

  4. Confirm scheduler active:

    /system scheduler print

Troubleshooting​

IssueCauseSolution
All vendos offlineNetwork connectivity lostPing vendo: ping 10.1.1.2; check cable connections
API key rejectedInvalid credentialsVerify key matches vendo provider documentation
No visible restartAPI endpoint wrong or firmware outdatedCheck vendo firmware version; test endpoint manually with curl
Restart doesn't completeVendo hanging during bootWait 2 minutes; manually power-cycle if stuck
Scheduler doesn't triggerTime/date mismatchCheck: /system clock print; sync time if needed

Advanced Options​

Add staggered delays to prevent network surge:​

:foreach vendo in=$vendos do={ 
    :delay 2s;
    /do {
        /tool fetch http-method=post http-header-field="X-TOKEN: $apiKey" \
            url="http://$vendo/admin/api/restartSystem"
    } on-error={
        :log error "FAILED: $vendo";
    }
}

Retry failed vendos:​

:local vendos {"10.1.1.2"; "10.1.2.2"; "10.1.3.2";};
:local apiKey "pqa92jwxrk";

:foreach vendo in=$vendos do={ 
    :local success 0;
    :for attempt from=1 to=3 do={
        :if ($success = 0) do={
            /do {
                /tool fetch http-method=post http-header-field="X-TOKEN: $apiKey" \
                    url="http://$vendo/admin/api/restartSystem"
                :set success 1;
                :log warning "Restarted $vendo on attempt $attempt";
            } on-error={
                :log error "Attempt $attempt failed for $vendo";
                :delay 1s;
            }
        }
    }
}

Send Telegram alert before restart:​

:local botApi "123456789:ABCDefGHIJKlmNOpqrsTUvwxYZ";
:local chatId "-123456789";
:local message "⚠️ Vendo restart scheduled - Service down 1 minute";
/tool fetch url="https://api.telegram.org/bot$botApi/sendMessage?chat_id=$chatId&text=$message" keep-result=no;

:delay 2s;

# Then execute restart...
:foreach vendo in=$vendos do={ ... }

Restart specific vendo groups by time:​

# Group A - Restart every Monday
:local vendosA {"10.1.1.2"; "10.1.2.2";};

# Group B - Restart every Thursday  
:local vendosB {"10.1.3.2"; "10.1.4.2";};

# Distribute load across week

Log restart history to file:​

:local timestamp [/system clock get date];
/file print file="vendo-restart-log" >> "$timestamp: Vendo restart initiated\n";

:foreach vendo in=$vendos do={ 
    /do {
        /tool fetch http-method=post http-header-field="X-TOKEN: $apiKey" \
            url="http://$vendo/admin/api/restartSystem"
        /file print file="vendo-restart-log" >> "$timestamp: $vendo - SUCCESS\n";
    } on-error={
        /file print file="vendo-restart-log" >> "$timestamp: $vendo - FAILED\n";
    }
}

Email notification on failed restarts:​

:local failedCount 0;

:foreach vendo in=$vendos do={ 
    /do {
        /tool fetch http-method=post http-header-field="X-TOKEN: $apiKey" \
            url="http://$vendo/admin/api/restartSystem"
    } on-error={
        :set failedCount ($failedCount + 1);
    }
}

:if ($failedCount > 0) do={
    /tool e-mail send to="admin@example.com" subject="Vendo Restart Failed" \
        body="$failedCount vendos failed to restart"
}

Conditional restart based on time:​

:local currentDay [/system clock get day-of-week];

# Only restart on weekends
:if ($currentDay = "sat" || $currentDay = "sun") do={
    :foreach vendo in=$vendos do={ ... }
}

Best Practices​

  1. Schedule during low-traffic hours (2-6 AM recommended)
  2. Notify customers before scheduled maintenance
  3. Stagger restarts if many vendos to prevent network overload
  4. Test on one vendo first before deploying to all
  5. Keep API keys secure and rotate regularly
  6. Monitor logs for repeated failures
  7. Have manual backup if API unavailable

Completion​

✅ Vendo restart automation is ready!

Next steps:

  • Test script on one vendo
  • Verify restart completes successfully
  • Schedule weekly maintenance window
  • Set up Telegram/email alerts
  • Back up configuration: /system backup save
  • Document restart schedule for team