High performance and scalable multi-layer topology discovery systems and methods
High performance and scalable multi-layer topology discovery systems and methods provide awareness of what services are present across a multi-layer network. The present disclosure achieves a high level of performance in service discovery, and, in addition, provides a form of scalability in proportion to network size by distributing service observation across servers in a cluster. The present disclosure defines a concept of change-proportional online run-time efficiency and thus provides an optimal design. Further, the present disclosure achieves horizontal scale, leveraging multiple cores across multiple servers.
1. A non-transitory computer-readable medium having instructions stored thereon for programming a processing device to perform steps of:
obtaining incoming data on a per network element basis, from network elements in a multi-layer network having a plurality of network elements and having at least a server layer and a client layer that operates over the server layer;
structuring the incoming data for topology discovery into atoms, each atom is a unit of data representing a server-server relationship or a client-server relationship;
processing the atoms at various layers including the server layer and the client layer in parallel to determine topology changes of services in the multi-layer network, wherein the various layers include any of Layer 0, 1, 2, and/or 3, wherein the server-server relationship models connectivity within one of the any of Layer 0, 1, 2, and/or 3, and wherein the client-server relationship models connectivity between adjacent layers of the any of Layer 0, 1, 2, and/or 3; and
detecting the topology changes by comparing the processed atoms to existing atoms, wherein the atoms are utilized to detect a complete service, for the topology changes, through a connected trail of atoms and a client-server relationship on both ends of the connected trail of atoms.
2. The non-transitory computer-readable medium of claim 1 , wherein the instructions further program the processing device to perform steps of
publishing output responsive to changes based on the incoming data and based on the processing the atoms.
3. The non-transitory computer-readable medium of claim 2 , wherein an amount of the processing is proportional to ultimate output change.
4. The non-transitory computer-readable medium of claim 1 , wherein the instructions further program the processing device to perform steps of
performing the processing the atoms concurrently across the server layer and the client layer.
5. The non-transitory computer-readable medium of claim 1 , wherein the instructions further program the processing device to perform steps of
utilizing a data structure in the structuring to detect a service completion at a layer, wherein the data structure represents connections in a graph, supporting a constant time lookup.
6. The non-transitory computer-readable medium of claim 1 , wherein the incoming data is asynchronous from the network elements.
7. An apparatus comprising:
a processor and memory including instructions that, when executed, cause the processor to
obtain incoming data on a per network element basis, from network elements in a multi-layer network having a plurality of network elements and having at least a server layer and a client layer that operates over the server layer,
structure the incoming data for topology discovery into atoms, each atom is a unit of data representing a server-server relationship or a client-server relationship,
process the atoms at various layers including the server layer and the client layer in parallel to determine topology changes of services in the multi-layer network, wherein the various layers include any of Layer 0, 1, 2, and/or 3, wherein the server-server relationship models connectivity within one of the any of Layer 0, 1, 2, and/or 3, and wherein the client-server relationship models connectivity between adjacent layers of the any of Layer 0, 1, 2, and/or 3; and
detect the topology changes by comparing the processed atoms to existing atoms, wherein the atoms are utilized to detect a complete service, for the topology changes, through a connected trail of atoms and a client-server relationship on both ends of the connected trail of atoms.
8. The apparatus of claim 7 , wherein the instructions that, when executed, further cause the processor to
publish output responsive to changes based on the incoming data and based on the processing the atoms.
9. The apparatus of claim 8 , wherein an amount of the processing is proportional to ultimate output change.
10. The apparatus of claim 7 , wherein the instructions that, when executed, further cause the processor to
perform the processing the atoms concurrently across the server layer and the client layer.
11. The apparatus of claim 7 , wherein the instructions that, when executed, further cause the processor to
utilize a data structure in the structuring to detect a service completion at a layer, wherein the data structure represents connections in a graph, supporting a constant time lookup.
12. The apparatus of claim 7 , wherein the incoming data is asynchronous from the network elements.
13. A method comprising:
obtaining incoming data on a per network element basis, from network elements in a multi-layer network having a plurality of network elements and having at least a server layer and a client layer that operates over the server layer;
structuring the incoming data for topology discovery into atoms, each atom is a unit of data representing a server-server relationship or a client-server relationship;
processing the atoms at various layers including the server layer and the client layer in parallel to determine topology changes of services in the multi-layer network, wherein the various layers include any of Layer 0, 1, 2, and/or 3, wherein the server-server relationship models connectivity within one of the any of Layer 0, 1, 2, and/or 3, and wherein the client-server relationship models connectivity between adjacent layers of the any of Layer 0, 1, 2, and/or 3; and
detecting the topology changes by comparing the processed atoms to existing atoms, wherein the atoms are utilized to detect a complete service, for the topology changes, through a connected trail of atoms and a client-server relationship on both ends of the connected trail of atoms.
14. The method of claim 13 , further comprising
publishing output responsive to changes based on the incoming data and based on the processing the atoms.
15. The method of claim 14 , wherein an amount of the processing is proportional to ultimate output change.
16. The method of claim 13 , further comprising
performing the processing the atoms concurrently across the server layer and the client layer.
17. The method of claim 13 , further comprising
utilizing a data structure in the structuring to detect a service completion at a layer, wherein the data structure represents connections in a graph, supporting a constant time lookup.