Data analytics on internal state

Patent metadata

Field	Value
Publication number	US-2016373302-A1
Application number	US-201615165358-A
Country	US
Kind code	A1
Filing date	May 26, 2016
Priority date	Jun 22, 2015
Publication date	Dec 22, 2016
Grant date	—

Abstract

Various embodiments are described herein to track and/or update the state of components within a network element. One embodiment provides for a network management system comprising a collector node including an ingest gateway to receive configuration and status data of a set of network elements coupled to the collector node via the network, the collector node further to store the configuration and status data from the set of network elements in a distributed database; a search engine to locate and retrieve the configuration and status data of a specified subset of the network elements at a specified time period; and an interface server to receive a request for the configuration and status data from an external requester, the interface server to facilitate provision of the configuration and status data in response to the request.

First claim

What is claimed is:

1 . A network management system comprising:
a collector node including an ingest gateway to receive configuration state and operational status of a set of network elements coupled to the collector node via a network, the collector node further to store state and status data from the set of network elements in a distributed database; an interface server to receive a request for the state and status data from an external requester, the interface server to facilitate provision of the state and status data in response to the request.

2 . The system as in claim 1 , wherein each network element in the set of network element includes a state and status agent to continuously transmit state and status data over the network.

3 . The system as in claim 2 , wherein the state and status agent of each network element is to transmit state and status data including status of one or more agents of a control plane of the network element.

4 . The system as in claim 3 , wherein the one or more agents of the control plane include a data-link layer forwarding agent and a network layer forwarding agent.

5 . The system as in claim 4 , wherein the state and status data additionally includes a media access control (MAC) address table stored in memory of the network element.

6 . The system as in claim 5 , wherein the state and status data additionally includes an Internet Protocol (IP) address table stored in memory of the network element.

7 . The system as in claim 6 , wherein the MAC address table and the IP address table are stored in a region of shared memory in the network element.

8 . The system as in claim 1 , wherein the state and status data additionally includes a value of one or more counters of each of an ingress interface and an egress interface of one or more network elements.

9 . The system as in claim 1 , additionally comprising an analysis agent to request the state and status data for a set of network elements from the interface server and analyze the state and status data of one or more network elements in the set of network elements.

10 . The system as in claim 9 , wherein the analysis agent is to request the configuration and status data at the specified time period encompassing a network failure event.

11 . The system as in claim 10 , wherein the interface server is configured to query a search engine to retrieve the requested state and status data from the distributed database and facilitate transfer of the requested configuration and status data to the analysis agent.

12 . The system as in claim 10 , wherein the analysis agent couples with a testing platform configured to replay events represented by the state and status data, the events including operations of the set of network elements in the time period prior to the network failure event.

13 . The system as in claim 12 , wherein the testing agent couples with one or more testing nodes, the testing nodes to receive the state and status data of the set of network elements to replay the events represented by the configuration and status data.

14 . The system as in claim 13 , wherein the one or more testing nodes include network elements physically coupled to the testing agent.

15 . The system as in claim 13 , wherein the one or more testing nodes include virtual network elements logically coupled to the testing agent.

16 . The system as in claim 1 , additionally comprising a monitoring agent to continuously monitor the state and status data of the set of network elements to predict a future network failure.

17 . The system as in claim 16 , wherein the monitoring agent is to predict the future network failure by monitoring a link status of one or more ports in the set of network elements.

18 . The system as in claim 16 , wherein the monitoring agent is to predict the future network failure by monitoring a state data rate of an interface in the set of network elements.

19 . The system as in claim 16 , wherein the monitoring agent is configured to monitor hardware memory errors of the various hardware components of the network elements.

20 . The system as in claim 16 , wherein the monitoring agent is to monitor a routing protocol state of the set of network elements.

21 . A non-transitory machine-readable medium storing instructions which, when executed by one or more processors, cause the one or more processors to perform operations including:
receiving configuration and status data associated with a first set of network elements, the configuration and status data associated with a time period encompassing a network failure involving one or more network elements in the set of network elements; applying initial configuration state and operational status associated with the set of network elements to a second set of network elements; and sequentially applying recorded configuration and status updates to the second set of network elements to determine a cause of the network failure.

22 . The medium of claim 21 , the operations additionally comprising querying an interface server of a network management system to request the state and status data and receiving the configuration and status data in response to the request.

23 . The medium of claim 21 , wherein the second set of network elements includes one or more virtual network elements executing within one or more virtual machines on a networked host device.

24 . A non-transitory machine-readable medium storing instructions which, when executed by one or more processors, cause the one or more processors to perform operations including:
querying an interface server of a network management system to request a subscription to configuration state and operational status data associated with a set of network elements; monitoring state and status data associated with the set of network elements via the subscription; and predicting a future network failure based on the configuration state and operational status data associated with the set of network elements.

25 . The medium of claim 24 , wherein the monitoring includes monitoring one or more of control plane state, shared memory state, and data plane state of the set of network elements, wherein the data plane state includes one or more interface counters of a network element.

26 . The medium of claim 25 , wherein predicting the future network failure includes predicting a future thermal related failure based on a current operational status of a network element in the set of network elements.

27 . The medium of claim 25 , wherein predicting the future network failure includes predicting a future hardware component failure within the set of network elements based on a pattern of memory errors related to the hardware component.

28 . The medium of claim 25 , wherein predicting the future network failure includes:
retrieving state and status data for a network element via the interface server of the network management system; querying an error database for observed errors correlated with the state and status data; and predicting a set of potential failures for the network element based on the querying.

29 . The medium of claim 25 , wherein predicting the future network failure includes determining whether an interface failure on a network element reduces forwarding capacity to a level insufficient for a peak observed networ

Assignees

• Arista Networks Inc

Inventors

• Sigoure Benoit

Classifications

• H04L41/0863

  by rolling back to previous configuration versions · CPC title

• H04L41/147

  for predicting network behaviour · CPC title

• H04L41/046

  comprising network management agents or mobile agents therefor · 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

• H04L41/0856 Primary

  by backing up or archiving configuration information · CPC title