We are experiencing a recurring issue where iOS devices using the 3CX mobile app stop ringing entirely until the 3CX services or PBX are restarted.
This appears to be a push delivery failure on the PBX rather than a provisioning or client configuration issue.
Environment
Symptoms
After some period of uptime:
However:
Event Log Evidence
Observed repeatedly in Event Log:
Deleting expired PUSH subscription of extension xxxx, device iPhone xx Pro
and:
Event ID 30035
Source: 3CX Event Notification Manager
The following service(s) were interrupted:
3CXGatewayService
3CXSystemService01
After restarting services, push resumes normally.
Observations
This appears consistent with push delivery subsystem failure or push subscription renewal failure.
The PBX continues processing SIP calls normally via Twilio trunk, but push notifications stop waking mobile clients.
This does not appear to be related to SIP trunk registration or inbound routing, as calls are successfully delivered to the PBX.
Questions
Additional context
Any insight into preventing push delivery from silently stopping would be appreciated.
This appears to be a push delivery failure on the PBX rather than a provisioning or client configuration issue.
Environment
- 3CX Version: V20 Enterprise Annual
- OS: Debian-based Linux VM (Hostinger VPS)
- Deployment: self-hosted
- SIP trunk: Twilio Elastic SIP Trunk
- Transport: Tunnel enabled for mobile clients
- Clients: multiple iPhones (13 Pro Max, 15 Pro, 16 Pro Max)
- Provisioning: QR provisioning
- Firewall: outbound unrestricted, DNS stable
Symptoms
After some period of uptime:
- Incoming calls do not ring on iPhone
- No CallKit UI appears
- Calls appear as missed in the 3CX app afterward
- Extension provisioning and push settings are correct
- Accept PUSH notifications enabled
- iOS permissions verified
- Reinstalling or reprovisioning mobile apps does not resolve
However:
- Restarting 3CX services or rebooting the PBX immediately restores normal behavior
- Push notifications resume normally after restart
- iPhones ring correctly again
Event Log Evidence
Observed repeatedly in Event Log:
Deleting expired PUSH subscription of extension xxxx, device iPhone xx Pro
and:
Event ID 30035
Source: 3CX Event Notification Manager
The following service(s) were interrupted:
3CXGatewayService
3CXSystemService01
After restarting services, push resumes normally.
Observations
This appears consistent with push delivery subsystem failure or push subscription renewal failure.
The PBX continues processing SIP calls normally via Twilio trunk, but push notifications stop waking mobile clients.
This does not appear to be related to SIP trunk registration or inbound routing, as calls are successfully delivered to the PBX.
Questions
- Is there a known issue in V20 Linux where GatewayService or Event Notification Manager stops delivering push notifications after extended uptime?
- Is there a supported way to restart only the push subsystem without restarting the entire PBX?
- Are expired push subscriptions expected to renew automatically, or do they require service restart?
- Are there recommended diagnostics to verify APNS connectivity from the PBX?
- Could Twilio trunk keepalive behavior or NAT state affect push delivery stability?
Additional context
- Issue affects multiple iOS devices on different networks
- Issue resolves immediately upon PBX service restart
- Strong indication of server-side push subsystem interruption
Any insight into preventing push delivery from silently stopping would be appreciated.
