Multi-directional fault detection system
View Patent ↗An apparatus, program product and method checks for nodal faults in a group of nodes comprising a center node and all adjacent nodes. The center node concurrently communicates with the immediately adjacent nodes in three dimensions. The communications are analyzed to determine a presence of a faulty node or connection.
1. A method for determining a nodal fault within a parallel processing system having a plurality of nodes, the method comprising:
determining a group comprised of the plurality of nodes;
designating one node of the group;
causing the designated node to communicate with adjacent nodes also within the group;
determining from the communications between the designated node and the adjacent nodes within the group the nodal fault relating to at least one of the plurality of nodes; and
after communications in the group have completed, creating another group and causing each node of the another group to communicate with another designated node.
2. The method of claim 1 , wherein designating the one node further comprises designating a center node of the group.
3. The method of claim 1 , wherein causing the designated node to communicate with adjacent nodes further comprises causing the designated and adjacent nodes to communicate concurrently.
4. The method of claim 1 , wherein causing the designated node to communicate with adjacent nodes further comprises determining if a first node in the group is adjacent to the designated node.
5. The method of claim 1 , further comprising servicing the nodal fault.
6. The method of claim 1 , wherein determining the nodal fault includes determining a fault associated with at least one of software, a connection, and another hardware component.
7. The method of claim 1 , wherein determining the nodal fault includes determining a latency associated with a nodal communication.
8. The method of claim 1 , wherein determining the nodal fault includes determining a bandwidth associated with a nodal communication.
9. The method of claim 1 , further comprising storing an indication of the nodal fault.
10. The method of claim 1 , further comprising generating a status signal indicative of the communication between the designated node and adjacent nodes.