Obviation of Recovery of Data Store Consistency for Application I/O Errors

• US 20120144233A1
• Filed: 12/01/2010
• Published: 06/07/2012
• Est. Priority Date: 12/01/2010
• Status: Active Grant

• Current Assignee: Kyndryl Incorporated
• Original Assignee: International Business Machines Corporation
• Inventors: Jaehde, Angela A., Krishnasamy, Somasundaram, Griffith, Douglas J., Schlachter, Stephen A.
• Granted Patent: US 8,713,362 B2
• Time in Patent Office: Days
• US Class Current: 714/13
• CPC Class Codes: G06F 11/1443 Transmit or communication errors
• CPC Class Codes: G06F 11/1482 by means of middleware or OS functionality
• CPC Class Codes: G06F 11/2007 using redundant communication media
• CPC Class Codes: G06F 11/202 where processing functionality is redundant redundant communication control functionality G06F11/2005, redundant storage control functionality G06F11/2089
• CPC Class Codes: G06F 2201/815 Virtual middleware or OS functionality using virtual machines to implement generic software techniques for error detection or fault masking G06F11/1484

Abstract

Embodiments comprise a plurality of computing devices that dynamically intercept process application I/O errors. Various embodiments comprise two or more computing devices, such as two or more servers, each having access to a shared data storage system. An application may be executing on the first computing device and performing an I/O operation when an I/O error occurs. The first computing device may intercept the I/O error, rather than passing it back to the application, and prevent the error from affecting the application. The first computing device may complete the I/O operation, and any other pending I/O operations not written to disk, via an alternate path, perform a checkpoint operation to capture the state of the set of processes associated with the application, and transfer the checkpoint image to the second computing device. The second computing device may resume operation of the application from the checkpoint image.

Claims

1. A method for handling an error of an input/output (I/O) operation, the method comprising:

a computer intercepting the error, wherein the I/O operation is via a first path to a shared storage system of data, wherein the intercepting comprises the computer preventing execution of code of the application in response to the error;

the computer completing the I/O operation via a second path to the shared storage;

the computer creating a checkpoint image of a set of processes that the application comprises, wherein the checkpoint image enables resumption of execution of the application, wherein the resumption of execution comprises starting execution at a point in the application subsequent to completion of the I/O operation; and

the computer transferring the checkpoint image to a second computer to enable the second computer to resume execution of the application via the checkpoint image.

2. The method of claim 1, further comprising:

monitoring, by a device driver of the first computer, the I/O operation to detect the error.

3. The method of claim 1, further comprising:

the computer halting execution of the application of the application to enable the creating the checkpoint image.

4. The method of claim 3, wherein the computer transferring the checkpoint image comprises transferring data of the checkpoint image via one of a network and a shared memory device, wherein the resumption of execution comprises executing additional I/O operations via a third path to the shared storage system.

5. The method of claim 3, further comprising:

the computer resuming the application prior to the creating the checkpoint image, wherein the computer resuming the application comprises the computer completing the I/O operation, wherein the second path comprises the network, wherein the network couples the computer with the second computer.

6. The method of claim 5, further comprising:

the computer caching a request of an additional I/O operation, wherein the computer comprises a server of a first cluster of servers, wherein the second computer comprises a server of a second cluster of servers, wherein the first path comprises a local connection of the server of the first cluster and the second path comprises a local connection of the server of the second cluster.

7. A computer program product comprising one or more computer-readable, tangible storage devices and computer-readable program instructions which are stored on the one or more storage devices and when executed by one or more processors, perform the method of claim 1.

8. A computer system comprising one or more processors, one or more computer-readable memories, one or more computer-readable, tangible storage devices and program instructions which are stored on the one or more storage devices for execution by the one or more processors via the one or more memories and when executed by the one or more processors perform the method of claim 1.

9. An apparatus, comprising:

input/output (I/O) hardware coupled to a storage device, wherein the I/O hardware enables performance of I/O operations for an application;

an application module to execute the application and generate a state of a set of processes for the application, wherein the application module is configured to perform an I/O operation for the application, wherein the I/O operation is via a first path to data of the storage device;

an error module to intercept an error of the I/O operation and prevent the error from causing the application module to execute code in response to receiving the error, wherein the application module is configured to complete the I/O operation via a second path to data of the storage device in response to the error module intercepting the error; and

a checkpoint module to create a checkpoint image of the state in response to completion of the I/O operation via the second path, wherein the checkpoint module is configured to transfer the checkpoint image to a second apparatus, wherein the checkpoint module is configured to create the checkpoint image in a manner which enables the second apparatus to resume execution of the application by starting execution at a point in the application subsequent to the completion of the I/O operation.

10. The apparatus of claim 9, further comprising a memory module to store instructions of the application module and a processor to execute the instructions.

11. The apparatus of claim 10, further comprising a first communication device which couples the apparatus to the storage device and enables the transfer of the data for the I/O operation.

12. The apparatus of claim 11, further comprising one of a network link and a shared memory module, wherein the network or the shared memory module enables the transfer of the checkpoint image to the second apparatus, wherein the storage device comprises a locally connected storage device coupled directly with the apparatus, wherein the second path comprises a second storage device coupled directly with a third apparatus.

13. The apparatus of claim 11, further comprising a network link that couples the apparatus to the second apparatus via transport control protocol/Internet protocol (TCP/IP) to enable the transfer of the checkpoint image, wherein the data storage device comprises a component of a storage area network (SAN), wherein the I/O hardware comprises a first physical host bus adapter (HBA) which couples the SAN to the apparatus, wherein the SAN is coupled to the second apparatus via a second physical HBA.

14. The apparatus of claim 13, further comprising a multi-path I/O (MPIO) module which enables the application module to transfer data via the network link and enables the application module to transfer data to the SAN via the first path.

15. The apparatus of claim 14, further comprising a management module configured to manage operations of the application module.

16. The apparatus of claim 15, further comprising a virtual machine monitor (VMM) configured to manage execution of the management module and the application module executes in a first partition, wherein the MPIO module comprises a driver module of a kernel, wherein the MPIO module comprises the error module.

17. A server for processing an error of an input/output (I/O) operation, the server comprising:

one or more processors, one or more computer-readable memories and one or more computer-readable, tangible storage devices;

program instructions, stored on at least one of the one or more storage devices for execution by at least one of the one or more processors via at least one of the one or more memories, to enable an application of the server to perform the I/O operation with a storage subsystem via a first storage connection, wherein the application of the server comprises a set of processes;

program instructions, stored on at least one of the one or more storage devices for execution by at least one of the one or more processors via at least one of the one or more memories, to detect the error of the set and prevent the error from causing one or more processes of the set from executing code in response to the error;

program instructions, stored on at least one of the one or more storage devices for execution by at least one of the one or more processors via at least one of the one or more memories, to complete the I/O operation via a second storage connection and create a checkpoint image of the set upon the completion of the I/O operation, wherein the checkpoint image enables resumption of execution of the application by starting execution at a point in the application subsequent to the completion of the I/O operation; and

program instructions, stored on at least one of the one or more storage devices for execution by at least one of the one or more processors via at least one of the one or more memories, to transfer the checkpoint image to a second server to enable the second server to resume operation of the application, wherein the transference is via at least one of a shared memory connection and a network interface.

18. The server of claim 17 wherein:

the program instructions to process the error of the I/O operation detect the error via a multi-path input/output (MPIO) module, wherein the MPIO module enables I/O operations between the server and the storage subsystem via the first storage connection and the second storage connection, wherein the second storage connection comprises a connection between a third server and the storage subsystem.

19. The server of claim 18, wherein:

the program instructions to process the error of the I/O operation complete the I/O operation and a plurality of other I/O operations via the second storage connection subsequent to detection of the error, wherein the program instructions to process the error of the I/O operation create the checkpoint image in response to completion of the plurality of other I/O operations.

20. The server of claim 17, wherein the application comprises an application of a virtual partition of the server, wherein the checkpoint image comprises an image of a resource group of the virtual partition.

21. A computer program product for handling an error of an input/output (I/O) operation, computer program product comprising:

one or more computer-readable, tangible storage devices;

program instructions, stored on at least one of the one or more storage devices, to intercept the error of the I/O operation of an application, wherein the I/O operation is via a first path from a first computing device to a shared storage system of data;

program instructions, stored on at least one of the one or more storage devices, to complete the I/O operation via a second path to the shared storage system of data;

program instructions, stored on at least one of the one or more storage devices, to create a checkpoint image of the application, wherein the checkpoint image comprises state of the application, wherein the program instructions to intercept the error comprise program instructions to prevent the first computing device from executing code of the application in response to the error, wherein the checkpoint image enables resuming execution of the application, wherein the resuming the execution comprises obviating initialization of the application; and

program instructions, stored on at least one of the one or more storage devices, to enable the generation of the checkpoint image in a second computing device.

22. The computer program product of claim 21, further comprising program instructions, stored on at least one of the one or more storage devices, to cache I/O operations of the application to enable completion of the I/O operations via the second path.

23. The computer program product of claim 21, further comprising program instructions, stored on at least one of the one or more storage devices, to complete the I/O operation via the second path prior to the resuming execution, wherein the second path comprises a link between the second computing device and the shared storage system of data.

24. The computer program product of claim 23, further comprising program instructions, stored on at least one of the one or more storage devices, to enable the first computing device to complete I/O operations of the application via a logical partition (LPAR), wherein the LPAR comprises instructions of a multi-path I/O (MPIO) module, wherein the instructions of the MPIO module enable the application to complete I/O operations of the via the first path and a third path, wherein the third path comprises a link between a third computing device and the shared storage system of data.