Managing failure behavior for computing nodes of provided computer networks

• US 9,736,016 B2
• Filed: 02/25/2015
• Issued: 08/15/2017
• Est. Priority Date: 10/27/2010
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Brandwine, Eric Jason
• Primary Examiner(s): Elliott, IV, Benjamin H
• Application Number:
  US14/631,675
• Publication Number:
  US 20150172104A1
• Time in Patent Office: 902 Days
• CPC Class Codes: H04L 41/064 involving time analysis
• CPC Class Codes: H04L 41/0654 using network fault recovery ring fault isolation or reconfiguration in loop networks without recovery actions by a network management system H04L12/437
• CPC Class Codes: H04L 41/0681 Configuration of triggering conditions
• CPC Class Codes: H04L 41/147 for predicting network behaviour
• CPC Class Codes: H04L 45/22 Alternate routing
• CPC Class Codes: H04L 45/28 using route fault recovery
• CPC Class Codes: H04L 69/40 for recovering from a failure of a protocol instance or entity, e.g. service redundancy protocols, protocol state redundancy or protocol service redirection management of faults, events, alarms or notifications in data switching networks H04L41/06

Abstract

Techniques are described for providing managed computer networks. In some situations, the techniques include managing communications for computing nodes of a managed computer network by using one or more particular computing nodes of the managed computer network that are configured to operate as intermediate destinations to handle at least some communications that are sent by and/or directed to one or more other computing nodes of the managed computer network. In addition, the techniques may include managing the communications in accordance with configured failure behavior specified for one or more computing nodes of the computer network, such as specified failure behavior for a computing node configured to operate as an intermediate destination that indicates how communications that would otherwise be routed via the intermediate destination computing node are to be handled if the intermediate destination computing node fails or is otherwise unavailable (e.g., to block or allow such communications).

Claims

1. A computer-implemented method comprising:

configuring, by one or more computing systems of a configurable network service that provides a virtual computer network, initial mapping information to use for forwarding communications for the virtual computer network, wherein the initial mapping information indicates that a first computing node of the virtual computer network is an intermediate destination for at least some communications sent during operation of the virtual computer network, and wherein the configuring includes configuring a manager that is associated with a second computing node of the virtual computer network to use the initial mapping information to manage communications sent by the second computing node;

receiving, by the configured manager, information regarding a first communication that is from the second computing node and to an intended final destination distinct from the first computing node;

determining, by the configured manager and based on the initial mapping information having an association of the intended final destination to the first computing node, to forward the first communication to the first computing node as an intermediate destination;

initiating, by the configured manager, forwarding the first communication to the first computing node;

updating, by the one or more computing systems while the operation of the virtual computer network is ongoing, and based on the first computing node becoming unavailable to act as the intermediate destination after the forwarding of the first communication, the initial mapping information to implement a failure behavior that is specified to be used when the first computing node becomes unavailable; and

managing, by the one or more computing systems and during further operation of the virtual computer network, a second sent communication for the virtual computer network by using the updated mapping information to determine a manner of handling the second sent communication based on unavailability of the first computing node.

2. The computer-implemented method of claim 1 further comprising, before the updating of the initial mapping information:

determining, by the one or more configured computing systems and based on the initial mapping information, to forward an indicated communication to an intended final destination without using the first computing node as an intermediate destination; and

initiating forwarding, over an underlying substrate network on which the virtual computer network is overlaid, the indicated communication to the intended final destination without providing the indicated communication to the first computing node.

3. The computer-implemented method of claim 1 wherein the updating of the initial mapping information includes configuring the manager to use the updated mapping information for communications sent by the second computing node after the updating, and wherein the method further comprises, after the updating of the initial mapping information:

receiving, by the configured manager, information regarding a third communication that is from the second computing node and to the intended final destination of the second communication;

determining, by the configured manager and based on the updated mapping information lacking the association of the intended final destination to the first computing node, an alternative manner of handling the third communication without forwarding the third communication to the first computing node as an intermediate destination; and

initiating, by the configured manager, handling the third communication in the determined alternative manner.

4. The computer-implemented method of claim 1 wherein the second sent communication would have been forwarded to the first computing node if the initial mapping information was in use, wherein the specified failure behavior indicates, while the first computing node is unavailable and for communications that would otherwise be forwarded to the first computing node as an intermediate destination, to forward those communications to intended final destinations of those communications, and wherein the managing of the second sent communication in the determined manner includes forwarding the second sent communication to an intended final destination for the second sent communication without forwarding the second sent communication to the first computing node.

5. The computer-implemented method of claim 1 wherein the specified failure behavior indicates to, while the first computing node is unavailable and for communications that would otherwise be forwarded to the first computing node as an intermediate destination, block those communications from being forwarded to intended final destinations of those communications, and wherein the managing of the second sent communication in the determined manner includes blocking the second sent communication from being provided to an intended final destination for the second sent communication and from being forwarded to the first computing node.

6. The computer-implemented method of claim 1 wherein the specified failure behavior indicates to, while the first computing node is unavailable and with respect to communications that would otherwise be forwarded to the first computing node as an intermediate destination, perform at least one action, wherein the managing of the second sent communication in the determined manner includes performing the at least one action without forwarding the second sent communication to the first computing node, and wherein the at least one action includes one or more actions from a group of actions that includes sending one or more notifications to a client of the configurable network service for whom the configurable network service provides the virtual computer network, removing one or more computing nodes from the virtual computer network, adding one or more computing nodes to the virtual computer network, executing one or more specified software programs, replacing the unavailable first computing node with another computing node that is configured to perform functionality of the first computing node, implementing a subset of the functionality of the first computing node for the communications that would otherwise be directed to the first computing node as an intermediate destination, or modifying the communications that would otherwise be directed to the first computing node as an intermediate destination to cause one or more networking devices in an underlying substrate network to determine how to further forward the modified communications.

7. The computer-implemented method of claim 1 further comprising, before the configuring of the initial mapping information:

receiving configuration information from a client of the configurable network service for the virtual computer network that specifies the failure behavior;

creating, by the configurable network service, the virtual computer network for the client based on the received configuration information; and

providing the virtual computer network for the client by overlaying the virtual computer network on a substrate network.

8. The computer-implemented method of claim 7 wherein the configuration information further includes a specified network topology for the virtual computer network that indicates the first computing node is located between a computing node that sends the second sent communication and an intended final destination for the second sent communication, wherein the configuring of the initial mapping information includes specifying the first computing node as the intermediate destination for the second sent communication based on the specified network topology, and wherein the providing of the virtual computer network is performed without physically implementing the specified network topology for the virtual computer network.

9. A non-transitory computer-readable medium having stored contents that cause one or more computing systems of a configurable network service to:

receive, by the configurable network service, configuration information for a virtual computer network provided by the configurable network service to a client, wherein the configuration information specifies failure behavior to implement if failure of a first computing node of the virtual computer network during operation of the virtual computer network causes the first computing node to be unavailable to act as an intermediate destination for communications, and wherein the configuration information further indicates one or more conditions during which the specified failure behavior is to be provided;

configure, by the configurable network service, one or more managers of the configurable network service to use initial mapping information for forwarding communications to the first computing node as the intermediate destination, and use the initial mapping information to forward one or more first sent communications to the first computing node;

modify, by the configurable network service and while the operation of the virtual computer network is ongoing and the first computing node is available, a current version of the mapping information to implement the specified failure behavior based on the one or more conditions being satisfied;

update, by the configurable network service while the operation of the virtual computer network is ongoing and based on the first computing node becoming unavailable to act as the intermediate destination, the initial mapping information to implement the specified failure behavior; and

manage, by the one or more managers and while further operation of the virtual computer network continues, a second sent communication for the virtual computer network by using the updated mapping information to determine to handle the second sent communication based on the specified failure behavior and without forwarding the second sent communication to the first computing node.

10. The non-transitory computer-readable medium of claim 9 wherein the second sent communication would have been forwarded to the first computing node if the initial mapping information was used, wherein the specified failure behavior indicates to, while the first computing node is unavailable and for communications that would otherwise be directed to the first computing node, forward those communications to intended final destinations of those communications, and wherein the managing of the second sent communication includes forwarding the second sent communication to an intended final destination for the second sent communication without forwarding the second sent communication to the first computing node.

11. The non-transitory computer-readable medium of claim 9 wherein the specified failure behavior indicates to, while the first computing node is unavailable and for communications that would otherwise be forwarded to the first computing node, block those communications from being forwarded to intended final destinations of those communications, and wherein the managing of the second sent communication includes blocking the second sent communication from being provided to an intended final destination for the second sent communication.

12. The non-transitory computer-readable medium of claim 9 wherein the stored contents include software instructions that, when executed, further cause the one or more computing systems to, before the updating of the initial mapping information:

provide the virtual computer network for the client by implementing functionality for a specified network topology of the virtual computer network indicated in the received configuration information but without physically implementing the specified network topology;

generate the initial mapping information based at least in part on the specified network topology;

determine to use the first computing node as the intermediate destination for an initial communication based on the initial mapping information; and

forward the initial communication to the first computing node based on the determining.

13. The non-transitory computer-readable medium of claim 9 wherein the stored contents include software instructions that, when executed, further cause the one or more computing systems to, after the managing of the second sent communication, receive information about updated failure behavior for the first computing node that corresponds to an alternative manner of handling communications while the first computing node is unavailable, and further update the mapping information to implement the updated failure behavior by causing the alternative manner to be used for handling one or more subsequent communications.

14. The non-transitory computer-readable medium of claim 9 wherein the stored contents include software instructions that, when executed, further cause the one or more computing systems to, after the managing of the second sent communication, determine that the first computing node has become available, further update the mapping information to terminate the specified failure behavior, and initiate forwarding a second third communication to the first computing node as an intermediate destination based on the further updated mapping information.

15. The non-transitory computer-readable medium of claim 14 wherein the first computing node is configured to manage forwarding of communications to their intended final destinations under additional specified conditions, and wherein the stored contents further cause the one or more computing systems to, after the forwarding of the third communication to the first computing node and if the additional specified conditions are satisfied, initiating further forwarding of the third communication from the first computing node to a final destination for the third communication.

16. The non-transitory computer-readable medium of claim 9 wherein the first computing node includes first and second logical network interfaces each having a distinct virtual network address for the virtual computer network, and wherein the specified failure behavior for the first computing node further indicates to logically interconnect the first and second logical network interfaces while the first computing node is unavailable.

17. The non-transitory computer-readable medium of claim 9 wherein the initial mapping information and the updated mapping information include associations of destination virtual network addresses for communications to indications of corresponding location indicators to which to forward those communications.

18. A non-transitory computer-readable medium having stored contents that cause one or more computing systems of a configurable network service to:

receive, by the configurable network service, configuration information for a virtual computer network provided by the configurable network service to a client, wherein the configuration information specifies failure behavior to implement if failure of a first computing node of the virtual computer network during operation of the virtual computer network causes the first computing node to be unavailable to act as an intermediate destination for communications, wherein the virtual computer network further includes a second computing node having a second associated failure behavior to be used when the second computing node becomes unavailable during operation of the virtual computer network, and wherein the second associated failure behavior is distinct from the specified failure behavior for use with the first computing node;

configure, by the configurable network service, one or more managers of the configurable network service to use initial mapping information for forwarding communications to the first computing node as the intermediate destination, and use the initial mapping information to forward one or more first sent communications to the first computing node;

update, by the configurable network service while the operation of the virtual computer network is ongoing and based on the first computing node becoming unavailable to act as the intermediate destination, the initial mapping information to implement the specified failure behavior;

manage, by the one or more managers and while further operation of the virtual computer network continues, a second sent communication for the virtual computer network by using the updated mapping information to determine to handle the second sent communication based on the specified failure behavior and without forwarding the second sent communication to the first computing node; and

further update the mapping information to implement the second associated failure behavior while the operation of the virtual computer network is ongoing based on the second computing node becoming unavailable.

19. A system comprising:

one or more processors of one or more computing systems; and

one or more memories with stored instructions that, when executed by at least one of the one or more processors, cause the system to;

configure initial mapping information to use with communications for a virtual computer network being provided, wherein the initial mapping information corresponds to a first computing node of the virtual computer network being available to act as an intermediate destination for at least some communications during operation of the virtual computer network;

update, while the operation of the virtual computer network is ongoing and based on a failure of the first computing node during the ongoing operation, the initial mapping information to implement a failure behavior that is specified to be used if the first computing node becomes unavailable to act as the intermediate destination, wherein the specified failure behavior indicates to, while the first computing node is unavailable and for communications that would otherwise be forwarded to the first computing node, forward those communications to intended final destinations of those communications; and

manage, during further operation of the virtual computer network, a sent communication for the virtual computer network by using the updated mapping information to determine a manner of handling the sent communication, wherein the sent communication would have been forwarded to the first computing node if the initial mapping information was used, and wherein the managing of the sent communication includes forwarding the sent communication to an intended final destination for the sent communication without forwarding the sent communication to the first computing node.

20. The system of claim 19 wherein the stored instructions further cause the system to, after the managing of the sent communication, modify the specified failure behavior to, while the first computing node is unavailable and for communications that would otherwise be forwarded to the first computing node, block those communications from being forwarded to intended final destinations of those communications, and mange an additional sent communication by using the modified specified failure behavior to determine to block the additional sent communication from being provided to an intended final destination for the additional sent communication.

21. The system of claim 19 wherein the stored instructions are part of a configurable network service and further cause the system to, before the updating of the initial mapping information:

receive configuration information from a client of the configurable network service for the virtual computer network, wherein the configuration information specifies the failure behavior and further specifies a network topology of the virtual computer network that includes the first computing node being located between a computing node that sends the sent communication and an indicated final destination of the sent computing node;

provide the virtual computer network for the client by implementing functionality for the specified network topology without physically implementing the specified network topology;

generate the initial mapping information based at least in part on the specified network topology;

determine to use the first computing node as the intermediate destination for a first communication based on the initial mapping information; and

forward the first communication to the first computing node based on the determining.