Automatic classification of network devices in a network

Patent metadata

Field	Value
Publication number	US-11700182-B2
Application number	US-202217739879-A
Country	US
Kind code	B2
Filing date	May 9, 2022
Priority date	Sep 28, 2018
Publication date	Jul 11, 2023
Grant date	Jul 11, 2023

Abstract

The automatic classification of network devices in a network. Specifically, the disclosure entails the designation of network device roles to network devices, as well as the clustering of network devices into logical groups. The association of network devices with network device roles and logical groups may be contingent on the connections between the network devices and a set of network device classification heuristics.

First claim

What is claimed is:

1. A method for network classification, comprising:
obtaining network topology information for a network, the network topology information comprising a plurality of nodes representing network devices, wherein the nodes are interconnected by a plurality of edges representing connections between pairs of the network devices; and
assigning network device roles to at least a portion of the plurality of nodes based on at least a portion of the plurality of edges and a set of network device classification heuristics, wherein assigning the network device roles comprises:
identifying a first set of the nodes as top-of-rack switch (TOR) candidates based on the edges connected to each of the nodes in the first set;
identifying a second set of the nodes as server candidates based on the edges connected to each of the nodes in the second set;
tagging a first portion of the TOR candidates as TORs based on the edges connected to each of the TOR candidates;
tagging a portion of the server candidates as servers based on the edges connected to each of the server candidates.

2. The method of claim 1 , wherein identifying the TOR candidates comprises: for each node of the first set of nodes: identifying a set of connections associated with the node; seeking a multi-chassis link aggregation group (MLAG) connection from the set of connections; making a determination that at least one MLAG connection has been identified.

3. The method of claim 2 , identifying, from an untagged portion of the first set of the nodes, a third set of the nodes as additional TOR candidates based on the edges connected to each of the untagged nodes, wherein identifying the additional TOR candidates comprises: identifying a set of connections associated with each untagged node; seeking a server connection from the set of connections; and making a determination that at least one server connection has been identified.

4. The method of claim 1 , wherein identifying the server candidates comprises: for each node of the second set of nodes: identifying a set of connections associated with the node; identifying a cardinality for the set of connections; making a determination that the cardinality is less than a threshold cardinality value; and assigning, based on the determination and to the node, a network device role tag associated with a candidate for the second class of network devices.

5. The method of claim 1 , further comprising identifying rack logical groups by:
(a) gathering rack-unassigned TORs and the TOR candidates in a prospective rack pool;
(b) identifying, from the prospective rack pool, a most-connected node;
(c) identifying, from a remainder of the prospective rack pool, one or more partner nodes which share children with the most-connected node;
(d) assigning a TOR tag and a rack logical group tag to the most-connected node and the one or more partner nodes;
(e) identifying, from the server and the server candidates, children of the most-connected node and the one or more partner nodes; and
(f) assigning a server tag, the rack logical group tag and a flock logical group tag to the identified children.

6. The method of claim 5 , further comprising determining whether the prospective rack pool is empty and, if the prospective rack pool is not empty, repeating steps (a)-(f).

7. The method of claim 1 , further comprising partitioning the plurality of nodes into a set of management nodes and a set of non-management nodes, wherein the set of management nodes comprises the first set of the nodes.

8. The method of claim 1 , further comprising partitioning the plurality of nodes into a set of management nodes and a set of non-management nodes, wherein the set of non-management nodes comprises the second set of the nodes.

9. A non-transitory computer readable medium (CRM) comprising computer readable program code, which when executed by a computer processor, enables the computer processor to perform a method for network classification, the method comprising:
obtaining network topology information for a network, the network topology information comprising a plurality of nodes representing network devices, wherein the nodes are interconnected by a plurality of edges representing connections between pairs of the network devices; and
assigning network device roles to at least a portion of the plurality of nodes based on at least a portion of the plurality of edges and a set of network device classification heuristics, wherein assigning the network device roles comprises:
identifying a first set of the nodes as top-of-rack switch (TOR) candidates based on the edges connected to each of the nodes in the first set;
identifying a second set of the nodes as server candidates based on the edges connected to each of the nodes in the second set;
tagging a first portion of the TOR candidates as TORs based on the edges connected to each of the TOR candidates;
tagging a portion of the server candidates as servers based on the edges connected to each of the server candidates.

10. The non-transitory CRM of claim 9 , wherein identifying the TOR candidates comprises: for each node of the first set of nodes: identifying a set of connections associated with the node; seeking a multi-chassis link aggregation group (MLAG) connection from the set of connections; making a determination that at least one MLAG connection has been identified.

11. The non-transitory CRM of claim 10 , identifying, from an untagged portion of the first set of the nodes, a third set of the nodes as additional TOR candidates based on the edges connected to each of the untagged nodes, wherein identifying the additional TOR candidates comprises: identifying a set of connections associated with each untagged node; seeking a server connection from the set of connections; and making a determination that at least one server connection has been identified.

12. The non-transitory CRM of claim 9 , wherein identifying the server candidates comprises: for each node of the second set of nodes: identifying a set of connections associated with the node; identifying a cardinality for the set of connections; making a determination that the cardinality is less than a threshold cardinality value; and assigning, based on the determination and to the node, a network device role tag associated with a candidate for the second class of network devices.

13. The non-transitory CRM of claim 9 , further comprising identifying rack logical groups by:
(a) gathering rack-unassigned TORs and the TOR candidates in a prospective rack pool;
(b) identifying, from the prospective rack pool, a most-connected node;
(c) identifying, from a remainder of the prospective rack pool, one or more partner nodes which share children with the most-connected node;
(d) assigning a TOR tag and a rack logical group tag to the most-connected node and the one or more partner nodes
(e) identifying, from the server and the server candidates, children of the most-connected node and the one or more partner nodes;
(f) assigning a server tag, the rack logical group tag and a flock logical group tag to the identified children; and
(g) determining whether the prospective rack pool is empty and, if the prospective rack pool is not empty, repeating steps (a)-(f).

14. The non-transitory CRM of claim 9 , further comprising: partitioning the plurality of nodes into a set of management nodes and a set of non-management nodes, wherein the set of management nodes comprises the first set of the nodes; and partitioning the plurality of nodes into a set of management nodes and a set of non-management nodes, wherein the set of non-management nodes comprises

Assignees

• Arista Networks Inc

Inventors

• Sirton Guy

Classifications

• H04L41/12 Primary

  Discovery or management of network topologies · CPC title

• G06N5/01

  Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound · CPC title

• H04L41/22

  comprising specially adapted graphical user interfaces [GUI] · CPC title

• G06N5/04

  Inference or reasoning models · CPC title

• H04L41/0893

  Assignment of logical groups to network elements · CPC title