QNX customers count on the high availability architecture and features of the QNX Neutrino RTOS. Whether the application is networking, safety-critical industrial automation, medical, military or automotive, QNX customers benefit from a high availability solution second to none in the embedded industry.

When it comes to achieving high availability (HA), the QNX® Neutrino® RTOS uses two simple principles:

• Minimize impact of software faults by containing failures to well defined software components
  
• Minimize mean time to repair (MTTR) by detecting faults and automating recovery without system reboots
  

These principles are embodied in the QNX Neutrino RTOS microkernel architecture. For example, if a device driver, protocol stack or application software experiences a problem, it can be immediately terminated and restored, often in a few milliseconds — no need to reboot. This limits the impact of the failure, allowing unaffected components to continue. The fine-grained approach to fault isolation and recovery is precisely why, compared to other operating systems, the QNX Neutrino RTOS provides a clear availability advantage.

Technology overview

• Microkernel architecture allows system services (such networking and file systems) and device drivers to be restarted automatically without affecting other software components
  
• Memory protection ensures errant software does not corrupt either the kernel or other applications
  
• POSIX process model enforces clear resource ownership and automatic cleanup by QNX Neutrino when processes terminate abnormally
  
• Built-in high availability framework
  
• Transparent distributed processing to quickly build redundant services
  
• Postmortem analysis with full memory dumps for postmortem debugging, allowing developers to quickly pinpoint memory violations and other error conditions.
  
• CPU time and memory guarantees (adaptive partitioning) to ensure critical software components have the resources needed to operate correctly.
  

High availability framework

Develop a sophisticated approach to failure detection and recovery, with the QNX® Neutrino's high availability framework. The high availability framework lets developers to quickly construct custom failure recovery scenarios and design your system to reconnect instantly and transparently to minimize downtime.

With the QNX high availability solution, developers can:

Build fault tolerant systems

• Automatically restart failed processes, without rebooting system
  
• Smart watchdog acts upon processes failures with support for heartbeat monitoring of system processes
  
• Mirror process ("Guardian") perpetually waits to take over monitoring role if necessary
  

Create customized failure recovery actions

• Applications select failure conditions (events) and specify actions to be performed when these occur
  

Automatically recover inter-process communication connections

• High availability connections provide automated re-connection after process failures