Predictive Self-Healing of Solaris™ 10 Operating System function
: [ Solaris Containers | Solaris ZFS | Predictive Self-healing | DTrace ]
A practiced home doctor for your servers
Around the world it is primarily the efforts of system administrators and field engineers that keep systems working as usual. But as systems grow, become more complex and more mission-critical, their burden of responsibility becomes very heavy.
System management and maintenance work requires highly skilled personnel. These are fast becoming a rare commodity, leading to extended service downtime, longer recovery times and increasing operational errors.
Fujitsu believes that server self-management is the way to go if we are to minimize sever failure, and reduce dependency on human intervention.
That is why Fujitsu SPARC Enterprise servers incorporate high-reliability and high-availability functions at the hardware-level that are intimately linked to the Predictive Self-Healing self-management functions in the Solaris 10 Operating System.
Predictive Self-Healing is a set of automated management functions that reduce the work of system management. They restart services swiftly, assist in pinpointing hardware errors and provide a rich array of analysis and problem isolation features. Two major new functions, Solaris Service Manager and Solaris Fault Manager have been included in the latest version.
- Solaris Service Manager unifies service management functions. This enables much swifter server start up and more rapid halt of the system following service failure.
- Solaris Fault Manager takes remedial action following server hardware error, providing the right prescriptive action to the system administrator
Solaris Service Manager
Solaris Service Manager reduces the complexity of systems administration. It significantly reduces server start-up time by starting parallel non-dependent services concurrently. Formerly, services were started one by one but now the Predictive Self-Healing functions know the interdependency of services. This means they can perform more efficient system start-up by starting independent services in parallel.
In addition, when a service fails, Solaris Service Manager automatically restarts all relevant and dependent services in the appropriate and correct order.
- Non-dependent services started in parallel not serially.
- Failed and dependent services are automatically restarted in the correct sequence
- Services can be monitored by their service-level
- Minimized server startup time
- Simplified server administration
- Simplified failure and service management
Solaris Fault Manager
Solaris Fault Manager is the diagnostic tool for server hardware. It finds and isolates problematic components allowing services to continue. It also writes a prescription for the system administrator on how to treat the hardware failure.
- Fails and isolates faulty components, collects and analyses failure information.
- Effect of the failure and prescriptive remedial action are reported to the administrator.
- Simplified server fault management
- Reduced error resolution times
Solaris Fault Manager with its server diagnosis function, further enhances SPARC Enterprise's availability profile, already well assured by high-quality hardware and server management functions.
Predictive Self-Healing wins InfoWorld 200 Innovator Award
DTrace and Predictive Self-Healing herald Sun's future (Info World)
Linux may be catching up to the Solaris of a few years ago, but ideas and technologies such as these prove that Solaris engineers are still a powerhouse of OS innovation.
The Solaris Services Manager facility requires configuration for services control.