System, computer program product and method in a node for optimizing the topology of a virtual ring based upon a TCP/IP network

• US 7,796,539 B2
• Filed: 10/31/2007
• Issued: 09/14/2010
• Est. Priority Date: 10/06/2005
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Denecheau, Lionel, Dispensa, Jean-Claude, Esteve, Denis
• Primary Examiner(s): Sheikh; Ayaz R
• Assistant Examiner(s): Ghafoerkhan; Faiyazkhan
• Application Number:
  US11/932,342
• Publication Number:
  US 20080056281A1
• Time in Patent Office: 1,049 Days
• Field of Search: None
• US Class Current: 370/254
• CPC Class Codes: H04L 12/42 Loop networks
• CPC Class Codes: H04L 69/16 Implementation or adaptation of Internet protocol [IP], of transmission control protocol [TCP] or of user datagram protocol [UDP]
• CPC Class Codes: H04L 69/161 Implementation details of TCP/IP or UDP/IP stack architecture; Specification of modified or new header fields
• CPC Class Codes: H04L 69/163 In-band adaptation of TCP data exchange; In-band control procedures
• CPC Class Codes: H04L 69/22 Parsing or analysis of headers

Abstract

In a network having nodes, N nodes are members of a virtual communications ring for multicast communication among the N Nodes, each of the N Nodes being associated with only an upstream and a downstream one of the other N nodes. One of the N nodes manages changes in ring membership, including requesting the others of the N nodes to participate in measuring distances to a N+1th, node responsive to receiving an insertion request message. The manager node inserts the N+1th node in the communications ring responsive to receiving replies, which includes inserting the N+1th node in a selected location between two selected ones of the N Nodes, such that the N nodes and the N+1th node are enabled to multicast messages around the ring. Due to the selected location of the N+1th node, communication distances for multicasted messages tend to be shorter than such distances that would otherwise occur.

Claims

1. A method in a network having nodes and a transport layer protocol providing end to end data transfer, wherein N of the nodes are members of a virtual communications ring for multicast communication among the N nodes, each of the N member nodes being logically associated with only a logically upstream one of the N nodes and a logically downstream one of the N nodes, and wherein for the communications ring one of the N nodes functions as a manager node for managing changes in ring membership, the method comprising:

receiving a first request from the manager node via the communication ring by one of the N member nodes, wherein the first request is received by the one node from the one node'"'"'s upstream node and is for signaling the one node to participate in insertion of a N+1th, non-member node in the communications ring;

forwarding, by the one node, the received first request to the one node'"'"'s downstream node in the communication ring; and

sending by the one node to the one node'"'"'s downstream node a reply for the manager node responsive to the received first request, wherein the receiving of the first request includes;

receiving a compute distance message from the manager node, wherein the compute distance message has an address of the N+1th node; and

wherein the method includes;

computing, by the one node, a distance between the one node and the N+1th node; and

wherein sending the reply includes;

sending a distance computed message for the manager node, wherein the distance computed message includes the computed distance;

wherein the reply provides the manager node information for enabling the N nodes and the N+1th node to multicast messages around the communication ring, so that the first request is passed around the ring back to the manager node and respective distances for the distance computer message are generated by the non-manager ones of the N nodes, wherein responsive to the replies the manager node selects a location for inserting the N+1th node as the N+1th member node between two selected ones of the N nodes in the communications ring such that communications distances for the multicasted messages tend to be shorter due to the selected location of the N+1th node than communications distances that would arise for other possible locations of the N+1th node in the communications ring.

2. The method of claim 1, wherein the method further includes:

receiving, by the one node, a second request from the manager node;

determining by the one node that the second request is a validation request;

forwarding, by the one node, the received validation request to the one node'"'"'s downstream node in the communication ring for validating topology of the ring by the validation request being passed around the ring so that the validation request is returned back to the manager node.

3. The method of claim 2, wherein the forwarding of the validation request is responsive to whether the validation request from the manager node is received by the one node via the one node'"'"'s upstream node.

4. The method of claim 2, wherein the receiving of the validation request by the one node includes receiving repeated instances of the validation request, and wherein the validation request includes indicia incremented by the managing node for the respective instances.

5. The method of claim 2, comprising the steps of:

timing, by the one node, respective intervals between instances of the validation request, wherein the timing is to confirm an alive state of the ring; and

detecting, by the one node, loss of the alive state of the ring responsive to expiration of a time limit for one of the intervals between instances of the validation request; and

sending, by the one node, a request to remove the one node from the ring responsive to detecting loss of the alive state.

6. The method of claim 1, wherein in the selecting of the location for inserting the N+1th node, the selecting is independent of distances between neighbouring ones of the N nodes.

7. The method of claim 1, wherein the reply from the one node includes identity of upstream and downstream neighbouring nodes, so that the manager node selects first and second nodes among the N nodes in the ring where between the manager node inserts the N+1th node, wherein the selecting is responsive to the identities of the upstream and downstream neighbouring nodes and the distances between respective nodes responding to the first request and the N+1th node and is independent of distances between neighbouring ones of the N nodes.

8. An apparatus comprising:

a first node coupled via a transport layer, network communication protocol in a virtual communications ring of nodes to provide end to end data transfer by multicast communication among the nodes, each of the member nodes being logically associated in the ring with only a logically upstream one of the nodes and a logically downstream one of the N nodes, wherein the first node includes a computer system coupled to a tangible, computer readable memory, wherein the computer system is programmed to;

receive a first request from an initiating one of the nodes via the communication ring, wherein the first request is received by the first node from the first node'"'"'s upstream node and is for signaling the first node to participate in insertion of a N+1th, non-member node in the communications ring;

forward the received first request to the first node'"'"'s downstream node in the communication ring; and

send to the first node'"'"'s downstream node a reply for the initiating node responsive to the received first request, wherein the receiving of the first request includes;

receiving a compute distance message from the initiating node, wherein the compute distance message has an address of the N+1th node; and

wherein the computer system is programmed to;

compute a distance between the one node and the N+1th node; and

wherein sending the reply includes;

sending a distance computed message for the initiating node, wherein the distance computed message includes the computed distance;

wherein the reply provides the initiating node information for enabling the N nodes and the N+1th node to multicast messages around the communication ring, so that the first request is passed around the ring back to the initiating node and respective distances for the distance computer message are generated by the non-initiating ones of the N nodes, wherein responsive to the replies the initiating node selects a location for inserting the N+1th node as the N+1th member node between two selected ones of the N nodes in the communications ring such that communications distances for the multicasted messages tend to be shorter due to the selected location of the N+1th node than communications distances that would arise for other possible locations of the N+1th node in the communications ring.

9. The apparatus of claim 8, wherein the computer system is programmed to receive a validation request from an initiating one of the N nodes, and to determine that the second request is a validation request;

and

forward the received validation request to the first node'"'"'s downstream node in the communication ring for validating topology of the ring by the validation request being passed around the ring so that the validation request is returned back to the initiating node.

10. The apparatus of claim 9, wherein the forwarding of the validation request is responsive to whether the validation request from the initiating node is received by the first node via the first node'"'"'s upstream node.

11. The apparatus of claim 9, wherein the computer system is programmed to:

time respective intervals between instances of the validation request, wherein the timing is to confirm an alive state of the ring;

detect loss of the alive state of the ring responsive to expiration of a time limit for one of the intervals between instances of the validation request; and

send a request to remove the one node from the ring responsive to detecting loss of the alive state.

12. The apparatus claim 8, wherein in the selecting of the location for inserting the N+1th node, the selecting is independent of distances between neighbouring ones of the N nodes.

13. The apparatus of claim 8, wherein the reply from the one node includes identity of upstream and downstream neighbouring nodes, so that the initiator node selects first and second nodes among the N nodes in the ring where between the initiator node inserts the N+1th node, wherein the selecting is responsive to the identities of the upstream and downstream neighbouring nodes and the distances between respective nodes responding to the first request and the N+1th node and is independent of distances between neighbouring ones of the N nodes.

14. A non-transitory computer-readable storage medium with an executable program stored thereon, wherein the program is for a computer system performing as a first node in a network having nodes and a transport layer protocol providing end to end data transfer, wherein N of the nodes are members of a virtual communications ring for multicast communication among the N nodes, each of the N member nodes being logically associated in the ring with only a logically upstream one of the N nodes and a logically downstream one of the N nodes, and wherein for the communications ring one of the N nodes functions as a manager node for managing changes in ring membership, wherein the programs instructs the computer system to perform the following:

receiving a first request from an initiating one of the nodes via the communication ring, wherein the first request is received by the first node from the first node'"'"'s upstream node and is for signaling the first node to participate in insertion of a N+1th, non-member node in the communications ring;

forwarding, by the first node, the received first request to the first node'"'"'s downstream node in the communication ring; and

sending to the first node'"'"'s downstream node a reply for the initiating node responsive to the received first request, wherein the receiving of the first request includes;

receiving a compute distance message from the initiating node, wherein the compute distance message has an address of the N+1th node; and

wherein the programs instructs the computer system to perform the following;

computing, responsive to the first request, a distance between the one node and the N+1th node; and

wherein the sending of the reply to the first node'"'"'s downstream node includes;

sending a distance computed message for the initiator node, wherein the distance computed message includes the computed distance;

wherein the reply provides the initiating node information for enabling the N nodes and the N+1th node to multicast messages around the communication ring, so that the first request is passed around the ring back to the initiating node and respective distances for the distance computer message are generated by the non-initiating ones of the N nodes, wherein responsive to the replies the initiating node to select a location for inserting the N+1th node as the N+1th member node between two selected ones of the N nodes in the communications ring such that communications distances for the multicasted messages tend to be shorter due to the selected location of the N+1th node than communications distances that would arise for other possible locations of the N+1th node in the communications ring.

15. The non-transitory computer-readable storage medium of claim 14, wherein the program instructs the computer system to perform the following:

receiving a second request from an initiating one of the N nodes, and for the first node determining that the second request is a validation request; and

forwarding the received validation request to the first node'"'"'s downstream node in the communication ring for validating topology of the ring by the validation request being passed around the ring so that the validation request is returned back to the initiating node.

16. The non-transitory computer-readable storage medium of claim 15, wherein the forwarding of the validation request is responsive to whether the validation request from the initiating node is received by the first node via the first node'"'"'s upstream node.

17. The non-transitory computer-readable storage medium of claim 14, wherein the receiving of the validation request by the first node includes receiving repeated instances of the validation request, and wherein the validation request includes indicia incremented by the initiating node for the respective instances, and wherein the forwarding by the first node of each instance of the validation request is responsive to verifying by the first node that the indicia has been incremented between instances.

18. The non-transitory computer-readable storage medium of claim 14, wherein the program instructs the computer system to perform the following:

timing respective intervals between instances of the validation request, wherein the timing is to confirm an alive state of the ring;

detecting, by the one node, loss of the alive state of the ring responsive to expiration of a time limit for one of the intervals between instances of the validation request; and

sending, by the one node, a request to remove the one node from the ring responsive to detecting loss of the alive state.

19. The non-transitory computer-readable storage medium of claim 14, wherein in the selecting of the location for inserting the N+1th node, the selecting is independent of distances between neighbouring ones of the N nodes.

20. The non-transitory computer-readable storage medium of claim 14, wherein the reply from the one node includes identity of upstream and downstream neighbouring nodes, so that the manager node selects first and second nodes among the N nodes in the ring where between the manager node inserts the N+1th node, wherein the selecting is responsive to the identities of the upstream and downstream neighbouring nodes and the distances between respective nodes responding to the first request and the N+1th node and is independent of distances between neighbouring ones of the N nodes.