Availability Guide for Problem Management

Auditing Systems for Fault Tolerance

Availability Guide for Problem Management–125509

7-10

Using Persistent Processes

Processes that only supply services to other processes but otherwise maintain no data of

their own need only to continue to execute. For such processes, it might be appropriate

simply to ensure that the process gets restarted whenever it stops. A monitor process that

periodically checks the process status can restart the process. Processes monitored in this

way are sometimes called persistent processes. These processes can employ either

immediate persistence or hot persistence, as follows:

•

Immediate persistence means that if a process stops, some other process can be

immediately notified so that the process can be immediately restarted.

•

Hot persistence means that a backup process is started in an alternate processor and

initialized. If the primary process fails, the backup process takes over in a pre-

initialized state.

Preventing External Data Communications Failures

You can use the Expand subsystem to connect your Tandem NonStop systems at

different locations to form a single network in which communications paths are

constantly available, even if a single line or component fails. If one path between nodes

fails, Expand automatically reroutes and (when appropriate) retransmits messages using

the next-best available path.

Another way to prevent outages caused by failures of external data communications is to

install redundant communications lines and devices.