Asynchronous communication in an unstable network

• US 8,755,397 B2
• Filed: 02/08/2012
• Issued: 06/17/2014
• Est. Priority Date: 06/03/2009
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Saha, Sanjib, Dar, Affan Arshad, Chen, Xin, Puttaswamaiah, Manohar Bangalore
• Primary Examiner(s): Jiang, Charles C
• Assistant Examiner(s): KAMARA, MOHAMED A
• Application Number:
  US13/368,705
• Publication Number:
  US 20120134370A1
• Time in Patent Office: 860 Days
• Field of Search: 370/389, 370/338, 370/412, 370/417, 370/392, 370/390, 370/252, 370/328, 370/254, 709/248, 709227-228
• US Class Current: 370/412
• CPC Class Codes: H04L 12/1895 for short real-time information, e.g. alarms, notifications, alerts, updates
• CPC Class Codes: H04L 12/56 Packet switching systems
• CPC Class Codes: H04L 41/12 Discovery or management of network topologies
• CPC Class Codes: H04L 51/043 using or handling presence information
• CPC Class Codes: H04L 51/48 Message addressing, e.g. address format or anonymous messages, aliases
• CPC Class Codes: H04L 61/2503 Translation of Internet protocol [IP] addresses
• CPC Class Codes: H04L 61/5076 Update or notification mechanisms, e.g. DynDNS
• CPC Class Codes: H04L 61/58 Caching of addresses or names

Abstract

Embodiments are directed to promptly reestablishing communication between nodes in a dynamic computer network and dynamically maintaining an address list in an unstable network. A computer system sends a message to other message queuing nodes in a network, where each node in the message queuing network includes a corresponding persistent unique global identifier. The computer system maintains a list of unique global identifiers and the current network addresses of those network nodes from which the message queuing node has received a message or to which the message queuing node has sent a message. The computer system goes offline for a period of time and upon coming back online, sends an announcement message to each node maintained in the list indicating that the message queuing node is ready for communication in the message queuing network, where each message includes the destination node'"'"'s globally unique identifier and the node'"'"'s current network address.

Claims

1. A computer-implemented method for promptly reestablishing communication between nodes in a dynamic computer network, the method being performed at a computer system comprising a message queuing node having a processor and a memory and that is disposed within a computer networking environment having a plurality of computing systems, the method comprising:

an act of a message queuing node sending a message to one or more other message queuing nodes in the dynamic computer network, each message queuing node in the message queuing network including a corresponding persistent unique global identifier, wherein the unique global identifier is unique to each particular message queuing node and is separate from and different than the node'"'"'s current network address and the node'"'"'s network hardware address;

an act of the message queuing node maintaining a list of unique global identifiers that identifies each and every network node from which the message queuing node has received a message or to which the message queuing node has sent a message over a given duration and to which an announcement message will be sent when the message queuing node comes back online after going off line for a period of time;

an act of the message queuing node going offline for the period of time; and

upon the message queuing node coming back online, an act of sending the announcement message to those nodes that are maintained in the list, the announcement message indicating that the message queuing node is ready for communication in the message queuing network, the announcement message including the node'"'"'s globally unique identifier and the node'"'"'s current network address.

2. The method of claim 1, further comprising:

an act of determining that a connection has been established between the message queuing node and at least one node from the list of nodes; and

based on the determination, an act of automatically resuming message transfer between the message queuing node and the at least one list node.

3. The method of claim 1, further comprising an act of broadcasting the announcement message containing the message queuing node'"'"'s identity and current network address to at least a plurality of the other nodes of the network.

4. The method of claim 1, wherein the unique global identifier is assigned to the node when the node is initially provisioned.

5. The method of claim 1, wherein each message queuing node'"'"'s unique global identifier is publicly available, such that applications can use the node'"'"'s unique global identifier to address messages to that node.

6. The method of claim 1, wherein when the message queuing node is offline, the message queuing node'"'"'s connection state is set to an offline state and the message queuing node'"'"'s network address value is set to null.

7. The method of claim 1, wherein the message queuing node comprises an address broadcast timer configured to initiate a broadcast of the node'"'"'s availability at predefined increments.

8. The method of claim 1, further comprising, upon the message queuing node coming back online, an act of the message queuing node sending any pending queued messages to nodes that have indicated they are available to receive messages.

9. The method of claim 1, wherein each node in the list includes the following information:

the unique global identifier of the destination node, the most recent network address of the destination node, session state indicating the current state of the network connection with the destination node, a collection of messages that are to be sent to the destination node, and timer that is started when the collection of messages has been successfully transmitted to the destination node.

10. The method of claim 9, wherein the timer is reset if an application adds new messages to be sent to the destination node.

11. The method of claim 10, further comprising:

an act of determining that the timer has expired; and

an act of removing the destination node from the list.

12. The method of claim 1, wherein the message queuing node listens for broadcast announcement messages on the network, such that when destination nodes come online with new network addresses, the message queuing node establishes a connection with the new network address of the destination node and begins transferring queued messages.

13. The method of claim 2, wherein communication between the message queuing node and the at least one list node is reinitiated automatically as soon as both are simultaneously online.

14. A computer storage device, comprising:

stored computer-executable instructions that, when executed by one or more processors of a message queuing node computing system, cause the message queuing node to perform a method for dynamically maintaining an address list in an unstable network, the method comprising;

an act of determining that a message queuing node has one or more outstanding queued messages that are to be sent to one or more other nodes;

an act of the message queuing node dynamically maintaining a globally unique identifier that is unique to that queued message node and is separate from the node'"'"'s network hardware address, and further maintaining a most recent network address for each of the nodes that are relevant to the message queuing node, wherein relevant nodes include all nodes for which the message queuing node has pending messages to be transmitted to the relevant node, or that have substantially recently communicated with the message queuing node;

an act of the message queuing node establishing a communication handshake sequence with the relevant nodes, such that if a relevant node'"'"'s network address changes, the message queuing node is apprised thereof and the destination node is configured not to accept any message that was received erroneously due to network address changes;

an act of determining that at least one of the relevant nodes is no longer using its last known network address; and

an act of storing any pending messages until the message queuing node receives an indication from the destination node indicating the node'"'"'s new network address and that the relevant node is available for communication.

15. The computer storage device of claim 14, the method further comprising an act of sending the stored pending messages to the relevant node using the node'"'"'s new network address.

16. The computer storage device of claim 14, wherein the relevant node'"'"'s indication comprises a broadcasted announcement message.

17. The computer storage device of claim 14, wherein the relevant node'"'"'s indication comprises a targeted announcement message directed at the message queuing node because the message queuing node had previously communicated with the relevant node.

18. The computer storage device of claim 14, wherein each node in the message queuing network is configured to dynamically maintain the most recent network addresses of the nodes with which each node communicates.

19. A computer system comprising the following:

one or more processors;

system memory;

one or more computer-readable storage media having stored thereon computer-executable instructions that, when executed by the one or more processors, causes the computing system to perform a method for promptly reestablishing communication between nodes in a dynamic computer network, the method comprising the following;

an act of a message queuing node sending a message to one or more other message queuing nodes in the dynamic computer network, each message queuing node in the message queuing network including a corresponding persistent unique global identifier, wherein the unique global identifier is unique to each particular message queuing node and is separate from and different than the node'"'"'s current network address and the node'"'"'s network hardware address;

an act of maintaining a list of unique global identifiers of those network nodes from which the message queuing node has received a message or to which the message queuing node has sent the message;

an act of the message queuing node going offline for a period of time; and

upon the message queuing node coming back online, an act of sending an announcement message to those nodes that are stored in the list, the announcement message indicating that the message queuing node is ready for communication in the message queuing network, the announcement message including the node'"'"'s globally unique identifier and the node'"'"'s current network address.

20. The system of claim 19, wherein the method further comprises:

an act of determining that a connection has been established between the message queuing node and at least one node from the list of nodes; and

based on the determination, an act of automatically resuming message transfer between the message queuing node and the at least one list node.