The physical layer clock protection inversion of Huawei OSN3500 protects the physical layer clocks of network elements and ensures that the physical layer clocks of each network element in the entire network are synchronized, which is very important to the service. To ensure that the physical layer clock protection can be reversed properly, the physical layer clock protection reversal must be tested. The physical layer clock includes the SDH clock, synchronized Ethernet clock, and OTN clock.

The failure recovery scenarios for OSN3500 equipment can be summarized into four categories:

Hardware failures such as line boards and crossover boards are the same as traditional SDH network disaster recovery mechanisms.

ASON intelligent software anomaly, service restart recovery.

Dual master control failure.

Database failure.

FailureScenarios	Consequences ofFailure	Recovery Methods
Branch board failure	Business interruption.	Replace the tributary board.
Line board failure	Business may be interrupted.	Reroute or optimize the service and replace the line board.
Double cross board failure	Business interruption.	Replace crossover boards.
DCCchannel failure	TElink degraded and cannot be used by intelligent services.	TroubleshootDCCchannel failure.
GSPsoftware abnormality	Software reset, service degradation, but no service interruption.	After the main control board restarts, the service resumes.
Single master failure, master and backup database synchronization	No effect.	The master and backup boards are reversed.
Single master failed, master and backup databases are not synchronized	Business may be interrupted after the master and backup are reversed.	If you can manually determine that the database is not synchronized, use the disaster recovery process disaster recovery solution; otherwise, the master and backup are reversed, the business is restarted and recovered, and the interrupted business is manually recovered.
Database failure	The business information on the node is lost, and the business may be interrupted.	If the database failure node can be started, then use the creation of intelligent business to rebuild the business, and if the business cannot be rebuilt, recover manually; if the database failure node cannot be started, usethe disaster recovery program shown inFigure1.
Dual master failure	The service will not be interrupted, but all the configuration data of the service and network elements on the node is lost.	After replacing the dual master control, recover according to the disaster recovery process.
Node Failure	Equivalent to dual master, dual crossover, and single board failure.	After replacing the device, follow the disaster recovery process to recover.