Failover mechanism for distributed process execution

• US 8,572,237 B2
• Filed: 12/16/2008
• Issued: 10/29/2013
• Est. Priority Date: 12/16/2008
• Status: Active Grant

• Current Assignee: SAP SE
• Original Assignee: SAP AG (SAP SE)
• Inventors: Ginkel, Thilo-Alexander, Klinker, Boris, Miltz, Matthias
• Primary Examiner(s): Lin, Kenny
• Assistant Examiner(s): Ali, Farhad
• Application Number:
  US12/336,500
• Publication Number:
  US 20100153481A1
• Time in Patent Office: 1,778 Days
• Field of Search: 709/224
• US Class Current: 709/224
• CPC Class Codes: G06F 15/16 Combinations of two or more digital computers each having at least an arithmetic unit, a program unit and a register, e.g. for a simultaneous processing of several programs coordinating program control therefor G06F9/52; in regulating and control system G05B
• CPC Class Codes: G06F 2209/505 Clust
• CPC Class Codes: G06F 9/5077 Logical partitioning of resources; Management or configuration of virtualized resources specific details on emulation or internal functioning of virtual machines G06F9/455

Abstract

Methods and apparatus, including computer program products, are provided for distributed process execution. In one aspect, the computer-implemented method may provide an indication announcing that a structural change to the cluster will occur. The indication may enable at least one node of the cluster to respond with an existing view. The existing view may include a first set of authoritative data representing existing processes handled at the at least one node. The method may also include providing, to at least one of the nodes, an index representing a subsequent view of the cluster after the structural change. The subsequent view may include a second set of authoritative data representing processes handled at the at least one node after the structural change. The method may also activate the subsequent view of the cluster. Related apparatus, systems, methods, and articles are also described.

Claims

1. A computer-implemented method comprising:

maintaining a state of a cluster comprising at least a plurality of nodes including a plurality of processors, a controller of the cluster, and a database, wherein a structural change is caused by at least one of adding a first node to the cluster and dropping a second node from the cluster, the structural change causing a change in state, wherein maintaining the state further comprises;

providing an indication announcing that the structural change to the cluster will occur, wherein the indication is provided by a node initiating the structural change, the indication enabling at least one node of the cluster to respond, after receiving a communication from the node initiating the structural change, with an existing view representing a topology defining the plurality of nodes from the perspective of the at least one node, the topology of defining an ordering of the plurality of nodes according to a sorting criterion, the existing view including a first set of authoritative data representing existing processes executed at the at least one node and processes for which the at least one node owns an authoritative index for but will not be authoritative subsequent to the structural change, wherein each node in the cluster independently determines the processes for which the node will not be authoritative for subsequent to the structural change;

providing, by the node initiating the structural change, to the plurality of nodes in the cluster an index representing a subsequent view of the cluster after the structural change, the subsequent view including a second set of authoritative data representing processes executed at the at least one node after the structural change, wherein each node in the plurality of nodes receiving the index determines at least one process, for which the node will be authoritative subsequent to the structural change; and

activating, by the node initiating the structural change, based on the index, the subsequent view of the cluster, when the provided index is received at the at least one node;

wherein, based on the activating, each node in the plurality of nodes, based on the index,

generates an index of processes that will be executed on the node;

generates a new cluster topology and provides the new cluster topology to the plurality of nodes in the cluster; and

removes at least one process identified in a previous index and the previous index associated with the node from the generated index of processes.

2. The computer-implemented method of claim 1 further comprising:

determining, by the first node joining the cluster, whether there is a cluster master assigned to the cluster.

3. The computer-implemented method of claim 2 further comprising:

assigning, based on the results of determining, the first node as the cluster master.

4. The computer-implemented method of claim 1 further comprising:

registering the first node joining the cluster with a cluster master assigned to the cluster.

5. The computer-implemented method of claim 1 further comprising:

using, as the authoritative data, a reclaim set and a relay set.

6. A non-transitory computer-readable medium including program code configured to provide operations comprising:

maintaining a state of a cluster comprising at least a plurality of nodes including a plurality of processors, a controller of the cluster, and a database, wherein a structural change is caused by at least one of adding a first node to the cluster and dropping a second node from the cluster, the structural change causing a change in state, wherein maintaining the state further comprises;

providing an indication announcing that the structural change to the cluster will occur, wherein the indication is provided by a node initiating the structural change, the indication enabling at least one node of the cluster to respond, after receiving a communication from the node initiating the structural change, with an existing view representing a topology defining the plurality of nodes from the perspective of the at least one node, the topology of defining an ordering of the plurality of nodes according to a sorting criterion, the existing view including a first set of authoritative data representing existing processes executed at the at least one node and processes for which the at least one node owns an authoritative index for but will not be authoritative subsequent to the structural change, wherein each node in the cluster independently determines the processes for which the node will not be authoritative for subsequent to the structural change;

providing, by the node initiating the structural change, to the plurality of nodes in the cluster an index representing a subsequent view of the cluster after the structural change, the subsequent view including a second set of authoritative data representing processes executed at the at least one node after the structural change, wherein each node in the plurality of nodes receiving the index determines at least one process, for which the node will be authoritative subsequent to the structural change; and

activating, by the node initiating the structural change, based on the index, the subsequent view of the cluster, when the provided index is received at the at least one node;

wherein, based on the activating, each node in the plurality of nodes, based on the index,

generates an index of processes that will be executed on the node;

generates a new cluster topology and provides the new cluster topology to the plurality of nodes in the cluster; and

removes at least one process identified in a previous index and the previous index associated with the node from the generated index of processes.

7. The non-transitory computer-readable medium of claim 6 further comprising:

determining, by the first node joining the cluster, whether there is a cluster master assigned to the cluster.

8. The non-transitory computer-readable medium of claim 7 further comprising:

assigning, based on the results of determining, the first node as the cluster master.

9. The non-transitory computer-readable medium of claim 6 further comprising:

registering the first node joining the cluster with a cluster master assigned to the cluster.

10. The non-transitory computer-readable medium of claim 6 further comprising:

using, as the authoritative data, a reclaim set and a relay set.

11. A system comprising:

a processor; and

a memory, the processor and memory configured to provide a method comprising;

maintaining a state of a cluster comprising at least a plurality of nodes including a plurality of processors, a controller of the cluster, and a database, wherein a structural change is caused by at least one of adding a first node to the cluster and dropping a second node from the cluster, the structural change causing a change in state, wherein maintaining the state further comprises;

providing an indication announcing that the structural change to the cluster will occur, wherein the indication is provided by a node initiating the structural change, the indication enabling at least one node of the cluster to respond, after receiving a communication from the node initiating the structural change, with an existing view representing a topology defining the plurality of nodes from the perspective of the at least one node, the topology of defining an ordering of the plurality of nodes according to a sorting criterion, the existing view including a first set of authoritative data representing existing processes executed at the at least one node and processes for which the at least one node owns an authoritative index for but will not be authoritative subsequent to the structural change, wherein each node in the cluster independently determines the processes for which the node will not be authoritative for subsequent to the structural change;

providing, by the node initiating the structural change, to the plurality of nodes in the cluster an index representing a subsequent view of the cluster after the structural change, the subsequent view including a second set of authoritative data representing processes executed at the at least one node after the structural change, wherein each node in the plurality of nodes receiving the index determines at least one process, for which the node will be authoritative subsequent to the structural change; and

activating, by the node initiating the structural change, based on the index, the subsequent view of the cluster, when the provided index is received at the at least one node;

wherein, based on the activating, each node in the plurality of nodes, based on the index,

generates an index of processes that will be executed on the node;

generates a new cluster topology and provides the new cluster topology to the plurality of nodes in the cluster; and

removes at least one process identified in a previous index and the previous index associated with the node from the generated index of processes.

12. The system of claim 11 further comprising:

determining, by the first node joining the cluster, whether there is a cluster master assigned to the cluster.

13. The system of claim 12 further comprising:

assigning, based on the results of determining, the first node as the cluster master.

14. The system of claim 11 further comprising:

registering the first node joining the cluster with a cluster master assigned to the cluster.

15. The system of claim 11 further comprising:

using, as the authoritative data, a reclaim set and a relay set.