Self-healing hierarchical network management system, and methods and apparatus therefor

Abstract

A hierarchical network management system (NMS) in which a plurality of NMS managers, each responsible for different portions or aggregations of a communications network, are logically arranged in a tree structure. The NMS managers are further organized into various sub-groups. The NMS managers within each sub-group monitor the status of one another in order to detect when one of them is no longer operational. If this happens, the remaining operational NMS managers of the sub-group collectively elect one of them to assume the responsibility of the non-operational NMS manager. The NMS is thus “self-healing” in the sense that one NMS manager can dynamically, without operator intervention, assume the responsibilities for another NMS manager.

1 . A method for managing a network, comprising: 
 arranging a plurality of network management system (NMS) managers in a hierarchy, said hierarchy having at least a root level and a leaf level, wherein each non-leaf level NMS manager supervises at least one child NMS manager and each leaf-level NMS manager supervises one or more network nodes;    determining when a given NMS manager ceases to operate; and    electing another NMS manager within said hierarchy to assume the responsibility of the non-operating NMS manager.    
     
     
         2 . The method according to  claim 1 , wherein, in the event a given NMS manager ceases to operate, the elected NMS manager is selected from a predetermined group of NMS managers within the hierarchy.  
     
     
         3 . The method according to  claim 2 , wherein the elected NMS manager is a sibling of the non-operating NMS manager.  
     
     
         4 . The method according to  claim 3 , wherein: 
 each leaf-level NMS manager receives state information pertaining to network elements under its supervision; and    each non-leaf level NMS manager receives aggregated state information pertaining to the network elements which are supervised by NMS managers that are descendent from the non-leaf level NMS manager.    
     
     
         5 . The method according to  claim 4 , wherein each NMS manager is implemented as a Holistic NMS and wherein the role of each such NMS Manager is dynamically configurable.  
     
     
         6 . The method according to  claim 5 , wherein the role of the NMS Manager is based on a network address.  
     
     
         7 . The method according to  claim 4 , wherein each NMS manager is implemented as a Segregated NMS.  
     
     
         8 . The method according to  claim 4 , wherein each NMS manager receives and stores state information pertaining to the network elements supervised by sibling NMS managers.  
     
     
         9 . The method according to  claim 8 , wherein each NMS manager includes an event service in order to publish to the siblings thereof events pertaining to network changes of state.  
     
     
         10 . The method according to  claim 9 , wherein the events include at least one of performance, connection, fault and configuration events.  
     
     
         11 . The method according to  claim 8 , wherein, for each group of sibling NMS manager, only one NMS manager within the group aggregates state information pertaining to all network elements supervised by the group to the common parent NMS manager.  
     
     
         12 . The method according to  claim 3 , wherein the determination of the non-operating NMS manager includes establishing a heartbeat process between at least two NMS manager siblings.  
     
     
         13 . The method according to  claim 1 , wherein the election is based on pre-configuration.  
     
     
         14 . The method according to  claim 1 , wherein the election is based on an administrative weight assigned to each NMS manager.  
     
     
         15 . The method according to  claim 1 , wherein the election is based the load bearing capability of each NMS manager.  
     
     
         16 . The method according to  claim 1 , wherein the election is based on network size.  
     
     
         17 . The method according to  claim 3 , wherein, in the event of an election, each NMS manager assumes it is the winner unless it receives notice otherwise from one of its siblings.  
     
     
         18 . The method according to  claim 4 , wherein each NMS manager within said hierarchy stores state information pertaining to the network elements under its sphere of responsibility to an external database such that the elected NMS manager can retrieve the state information associated with the non-operating NMS manager.