Modular architecture for computer network operations

Patent metadata

Field	Value
Publication number	US-10693774-B1
Application number	US-201816162930-A
Country	US
Kind code	B1
Filing date	Oct 17, 2018
Priority date	May 9, 2016
Publication date	Jun 23, 2020
Grant date	Jun 23, 2020

Abstract

Disclosed embodiments include a method of operation of a distributed network system. The method includes nodes of the network system that send messages over a protocol-independent message bus, and other nodes that receive the messages. Content from the received messages can be stored in a database distributed among nodes of the network system. At least some of the content stored in the database is published. The published content can be accessed by one or more applications to perform one or more functions.

First claim

We claim:

1. A computer-implemented method for mitigating flooding in a network comprising:
segmenting one or more network nodes into one or more clusters of nodes, such that each cluster has at least one network node;
for each of the one or more clusters of nodes:
computing a spanning tree for the cluster of nodes, each spanning tree comprising a logical loop-free topology that ensures that each message flooded to the cluster traverses the cluster's spanning tree only once;
flooding at least one message to the network, wherein each of the at least one messages follows the computed spanning tree loop-free topology for the cluster to which each of the at least one messages is sent; and
tracing a message path for each of the at least one messages to prevent each of the at least one messages from being sent more than once to the same cluster.

2. The computer-implemented method of claim 1 , wherein segmenting the one or more network nodes into the one or more clusters of nodes mitigates message flooding overhead by preventing each of the at least one messages from repeatedly traversing a same network segment.

3. The computer-implemented method of claim 2 , wherein computation of the spanning tree is distributed across the one or more network nodes.

4. The computer-implemented method of claim 1 , wherein message flooding occurs between a message source and a first cluster.

5. The computer-implemented method of claim 4 , wherein the message flooding occurs between the first cluster and a second cluster, such that the second cluster receives each of the at least one messages from the first cluster.

6. The computer-implemented method of claim 1 , wherein the network is organized in a substantially uniform topology.

7. The computer-implemented method of claim 6 , wherein the network with the substantially uniform topology comprises a datacenter.

8. The computer-implemented method of claim 1 , wherein tracing the message path for each of the at least one messages includes appending one or more portions of data to each of the at least one messages.

9. The computer-implemented method of claim 8 , wherein the appended one or more portions of data indicates which nodes or clusters were previously traversed by each of the at least one messages.

10. The computer-implemented method of claim 9 , wherein messages arriving at a specified cluster having the appended one or more portions of data indicating that the specified cluster has already been traversed are discarded.

11. The computer-implemented method of claim 8 , wherein the one or more portions of data are appended in a specified format defined by a message bus.

12. The computer-implemented method of claim 11 , wherein the specified format includes one or more custom fields for tracking the message path through the network.

13. A system comprising:
at least one physical processor;
physical memory comprising computer-executable instructions that, when executed by the physical processor, cause the physical processor to:
segment one or more network nodes in a network into one or more clusters of nodes, such that each cluster has at least one network node;
for each of the one or more clusters of nodes:
compute a spanning tree for the cluster of nodes, each spanning tree comprising a logical loop-free topology that ensures that each message flooded to the cluster traverses the cluster's spanning tree only once;
flood at least one message to the network, wherein each of the at least one messages follows the computed spanning tree loop-free topology for the cluster to which each of the at least one messages is sent; and
trace a message path for each of the at least one messages to prevent each of the at least one messages from being sent more than once to the same cluster.

14. The system of claim 13 , wherein the spanning trees are calculated using a spanning tree algorithm that is distributed across the one or more network nodes.

15. The system of claim 14 , wherein the distributed spanning tree algorithm runs using a hello protocol to compute the spanning tree for each cluster of nodes.

16. The system of claim 13 , wherein at least one specified network node is configured to drop packets that have either already traversed a specified node or have traversed the cluster of nodes to which the specified node belongs.

17. The system of claim 16 , wherein messages are appended with cluster identifiers as the messages traverse clusters in the network.

18. The system of claim 16 , wherein each of the at least one messages includes a hop count which allows the one or more network nodes to drop messages that are caught in transient routing loops.

19. The system of claim 13 , wherein a number of clusters in the network that include spanning trees is calculated to provide a specified level of message flooding scalability within the network.

20. A non-transitory computer-readable medium comprising one or more computer-executable instructions that, when executed by at least one processor of a computing device, cause the computing device to:
segment one or more network nodes in a network into one or more clusters of nodes, such that each cluster has at least one network node;
for each of the one or more clusters of nodes:
compute a spanning tree for the cluster of nodes, each spanning tree comprising a logical loop-free topology that ensures that each message flooded to the cluster traverses the cluster's spanning tree only once;
flood at least one message to the network, wherein each of the at least one messages follows the computed spanning tree loop-free topology for the cluster to which each of the at least one messages is sent; and
trace a message path for each of the at least one messages to prevent each of the at least one messages from being sent more than once to the same cluster.

Assignees

• Facebook Inc

Inventors

• Lapukhov Petr V
• Saab Paul M
• Baldonado Omar Castillo
• Williams James
• Somasundaram Manikandan

Classifications

• H04L41/12

  Discovery or management of network topologies · CPC title

• H04L45/484

  using multiple routing trees · CPC title

• H04L12/44 Primary

  Star or tree networks · CPC title

• H04L63/08

  for authentication of entities (cryptographic mechanisms or cryptographic arrangements for entity authentication H04L9/32) · CPC title

• H04L67/1097

  for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS] · CPC title