Fault tolerance solution for stateful applications

• US 9,110,864 B2
• Filed: 06/25/2013
• Issued: 08/18/2015
• Est. Priority Date: 06/25/2013
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Jamjoom, Hani T., Rajagopalan, Shriram, Williams, Daniel J.
• Primary Examiner(s): CHU, GABRIEL L
• Application Number:
  US13/925,872
• Publication Number:
  US 20140380087A1
• Time in Patent Office: 784 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06F 11/1484 involving virtual machines
• CPC Class Codes: G06F 11/2002 where interconnections or communication control functionality are redundant flexible arrangements for bus networks involving redundancy H04L12/40176
• CPC Class Codes: G06F 11/2028 eliminating a faulty processor or activating a spare
• CPC Class Codes: G06F 11/2038 with a single idle spare processing component

Abstract

A fault tolerance method and system for VMs on a cluster identifies a client state for each client session for those applications. The method replicates the client session onto a primary and a backup VM, and uses a network controller and orchestrator to direct network traffic to the primary VM and to periodically replicate the state onto the backup VM. In case of a VM failure, the method reroutes network traffic of states for which the failed VM serves as a primary to the corresponding backup, and replicates states without a backup after the failure onto another VM to create new backups. The method may be used as part of a method or system implementing the split/merge paradigm.

Claims

1. A method for providing fault tolerance on a virtual machine (VM) cluster, comprising:

maintaining a plurality of VMs in a VM cluster servicing a plurality of client sessions each having a network traffic flow directed to the VM cluster;

generating a primary client state and a backup client state for each client session according to a predefined criteria, wherein the primary client state and the backup client state are hosted on separate instances of the VMs in the VM cluster;

directing the network traffic flow of each of the client sessions to the VM hosting the primary client state of the client session;

detecting a failing VM in the VM cluster;

designating the backup client states of the primary client states hosted on the failing VM as new primary client states and directing the network traffic flow of the corresponding client sessions to the VMs hosting the new primary client states; and

generating a new backup client state for each of the backup client states hosted on the failing VM and a new backup for each of the new primary client states.

2. The method of claim 1, wherein the plurality of VMs in the VM cluster run a corresponding plurality of applications that service one or more of the plurality of client sessions and their respective client states.

3. The method of claim 2, wherein at least one of the plurality of applications is a session oriented network application or a network flow oriented application.

4. The method of claim 2, wherein at least one of the plurality of applications is a webserver or a middlebox application.

5. The method of claim 1, wherein the step of generating a backup client state is performed periodically.

6. The method of claim 1, wherein a frequency of performing the step of generating a client state and a number of backups is configurable.

7. The method of claim 1, wherein the predefined criteria includes a split/merge paradigm.

8. A system for providing fault tolerance on a virtual machine (VM) cluster, comprising:

a first computer having a processor, and a computer-readable storage device; and

a program embodied on the storage device for execution by the processor, the program having a plurality of program modules, including;

a maintaining module configured to maintain a plurality of VMs in a VM cluster servicing a plurality of client sessions each having a network traffic flow directed to the VM cluster;

a first generating module configured to generate a primary client state and a backup client state for each client session according to a predefined criteria, wherein the primary client state and the backup client state are hosted on separate instances of the VMs in the VM cluster;

a directing module configured to directing the network traffic flow of each of the client sessions to the VM hosting the primary client state of the client session;

a detecting module configured to detect a failing VM in the VM cluster;

a designating module configured to designate the backup client states of the primary client states hosted on the failing VM as new primary client states and to direct the network traffic flow of the corresponding client sessions to the VMs hosting the new primary client states; and

a second generating module configured to generate a new backup client state for each of the backup client states hosted on the failing VM and a new backup for each of the new primary client states.

9. The system of claim 8, wherein the plurality of VMs in the VM cluster run a corresponding plurality of applications that service one or more of the plurality of client sessions and their respective client states.

10. The system of claim 9, wherein at least one of the plurality of applications is a session oriented network application or a network flow oriented application.

11. The system of claim 9, wherein at least one of the plurality of applications is a webserver or a middlebox application.

12. The system of claim 8, wherein the step of generating a backup client state in the first generating module is performed periodically.

13. The system of claim 8, further comprising:

a plurality of additional computers in a network, each additional computer having a processor and a computer readable storage device configured to execute the program of the first computer, wherein each of the first computer and the additional computers maintain one or more VMs in the VM cluster.

14. The system of claim 8 further comprising a network controller and an orchestrator.

15. The system of claim 8 wherein the predefined criteria includes a split/merge paradigm.

16. A computer program product for providing fault tolerance on a virtual machine (VM) cluster, the computer program product comprising a non-transitory computer-readable storage medium having program code embodied therewith, the program code readable/executable by a first processor of a first computer to perform a method comprising:

maintaining a plurality of VMs, by the processor, in a VM cluster servicing a plurality of client sessions each having a network traffic flow directed to the VM cluster;

generating a primary client state and a backup client state, by the processor, for each client session according to a predefined criteria, wherein the primary client state and the backup client state are hosted on separate instances of the VMs in the VM cluster;

directing the network traffic flow of each of the client sessions, by the processor, to the VM hosting the primary client state of the client session;

detecting a failing VM in the VM cluster, by the processor;

designating the backup client states of the primary client states hosted on the failing VM as new primary client states, by the processor, and directing the network traffic flow of the corresponding client sessions, by the processor, to the VMs hosting the new primary client states; and

generating a new backup client state, by the processor, for each of the backup client states hosted on the failing VM and a new backup for each of the new primary client states.

17. The computer program product of claim 16, wherein the plurality of VMs in the VM cluster run a corresponding plurality of applications that service one or more client states and their respective client sessions.

18. The computer program product of claim 17, wherein at least one of the plurality of applications is a session oriented network application or a network flow oriented application.

19. The computer program product of claim 16, wherein a frequency of performing the step of generating a backup client state and a number of backups is configurable.

20. The computer program product of claim 16, wherein the predefined criteria includes a split/merge paradigm.