STORAGE MANAGEMENT SYSTEM WITH INTEGRATED CONTINUOUS DATA PROTECTION AND REMOTE COPY

Abstract

A storage control system is adapted to operate as a primary of a remote copy pair by communicating with a secondary of the remote copy pair, and comprises: an I/O receiving component for receiving read and write commands; a continuous data protection component operatively coupled to the I/O receiving component for maintaining a continuous data protection journal; the continuous data protection journal operable to retain write I/O data items for access by subsequent read commands and to roll back the data at the primary of the remote copy pair; a remote copy controller component adapted to read the write I/O data item from the continuous data protection journal; and a remote copy transmission component operatively coupled to the remote copy controller component to transmit the write I/O data item to the secondary of the remote copy pair.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments will now be described, by way of example only, with reference to the accompanying drawing figures, in which:

FIG. 1 shows in schematic form an arrangement of data structures.

FIG. 2 shows a format of a header of a record.

FIG. 3 shows in schematic form an arrangement of components; and

FIG. 4 shows in flowchart form one method or one logic arrangement.

DETAILED DESCRIPTION

Certain described embodiments are broadly directed to a CDP scheme which is tightly integrated with remote copy. This may be achieved as described below, with reference to the figures.

With respect to FIG. 1, the storage controller divides each logical volume into pages, for example, 4 KB each (which could represent eight 512-byte blocks.) Many applications generate I/O requests of this size, or a multiple thereof. When an application issues a write command 2, the controller appends a Record 4a, 4b, 4c, 4d containing the Page 6a, 6b, 6c, 6d to a sequential Journal 8 which is stored on one or more disk arrays. A single Journal 8 is used to store the writes to all logical volumes managed by a controller. Thus the Journal 8 automatically preserves all previous versions of each Page 6a, 6b, 6c, 6d, in sequence, so that they are available for a subsequent rollback. In one embodiment, updates to the Journal 8 may be accumulated in an underlying non-volatile write cache and destaged to RAID-5 arrays as full Strides in order to minimize the number of disk accesses. In a further embodiment, as the Journal 8 is always written sequentially, the contents of a Record 4a, 4b, 4c, 4d may readily be compressed to save storage space.

The controller also maintains a Directory 10 for each logical volume so that it can locate the contents of a Page in the Journal when it receives a read command. The Directory 10 is hardened on disk. It may also be cached in memory so that Pages 6a, 6b, 6c, 6d which are accessed frequently can be located quickly. The controller must update the Directory 10 for each write command 2 it receives.

There are many possible designs for the Directory 10. In the simplest case the Directory 8 is just a linear array of Pointers 12a, 12b, 12c, 12d to the current Record 4a, 4b, 4c, 4d for each Page 6a, 6b, 6c, 6d. A null Pointer 12a, 12b, 12c, 12d indicates that the corresponding Page 6a, 6b, 6c, 6d has never been written. More complex Directories could be designed with a time dimension in order to rapidly locate the contents of each Page at previous points in time.

With respect to FIG. 2, each Record 4a, 4b, 4c, 4d in the Journal 8 preferably contains a Header 50. The Header 50 may include the number of compressed bytes in the Page 52, an identifier for the logical volume 54, the starting LBA 56 and a Timestamp 58. Thus each record 4 is preferably self-describing and the Headers 50 can be used as a recovery mechanism to rebuild the Directory 10 by scanning the Journal 8 from the beginning. With the simple Directory 10 described above, the Header 50 may also include one or more backward Pointers 60 to Records 4a, 4b, 4c, 4d containing previous versions of the Page 6a, 6b, 6c, 6d. This allows the contents of a Page 6a, 6b, 6c, 6d to be rolled back to a previous point in time.

Initially the Journal 8 is empty and the Directory 10 contains an array of null Pointers 12a, 12b, 12c, 12d. When the controller receives a write command 2 for a single Page 6a, 6b, 6c, 6d, it appends a Record 4a, 4b, 4c, 4d to the Journal 8 and then it updates the corresponding Pointer 12a, 12b, 12c, 12d in the Directory 10. Longer writes cause the same action for each Page 6a, 6b, 6c, 6d. Unaligned or short writes are processed by reading a Page 6a, 6b, 6c, 6d, modifying one or more blocks and then writing the Page.

Reads 14 are processed by indexing down the Directory 10 to locate the corresponding Pointer 12a, 12b, 12c, 12d and then following it to a Record 4a, 4b, 4c, 4d in the Journal 8. When the data has been rolled back to a previous point in time it is necessary to follow the backward pointers 60 in each Header 50 to find the Record 4a, 4b, 4c, 4d with the correct Timestamp 58. Subsequently a new Directory could be constructed to restore full performance. This will take some time but it can be done concurrently while data access continues using the old Directory and the backward Pointers 60.

Eventually the Journal 8 will fill up and space must be freed by deleting some previous versions of the data. This is done by a background Garbage Collector which scans the Journal 8 from the beginning and deletes some previous Records 4a, 4b, 4c, 4d for each Page 6a, 6b, 6c, 6d according to a policy; in one exemplary embodiment, the policy might be “keep one version for each previous week up to a month”. The Garbage Collector copies each Record 4a, 4b, 4c, 4d that is retained to the first available empty slot in the Journal 8 in order to defragment it, and it updates the Directory 10 accordingly.

Remote copy is implemented by copying each new Record 4a, 4b, 4c, 4d from the primary Journal 8 to a secondary Journal 16 at the remote site. This can be done synchronously with each write command 2 before returning completion status to the host. Alternatively, it can be an asynchronous process which may run some Records behind when the link speed cannot keep up with the I/O rate. In both cases, a secondary Directory 18 is updated to point 20a, 20b, 20c, 20d to each record 22a, 22b, 22c, 22d which is received at the remote site. In some situations the controller may switch automatically from synchronous remote copy to asynchronous remote copy and vice versa, for example if the link fails and is later repaired.

If the controller has a cache then the functions described here may be implemented in a layer above the cache. This avoids the ordering issues potentially caused by fast write and allows the usual benefits of a read cache. A Freeze/Thaw protocol can be used to ensure consistency across multiple controllers by entering sync points into each Journal.

With respect to FIG. 3, there is shown a storage control system in which a primary controller 100 is adapted to operate as a primary of a remote copy pair by communicating with a secondary controller 150 of the remote copy pair. Primary controller 100 and secondary controller 150 thus form the controlling components of the remote copy pair.

Primary controller 100 comprises an I/O receiving component 104 for receiving read and write commands and a continuous data protection control component 106 operatively coupled to the I/O receiving component 104 for maintaining a continuous data protection journal 108 and a directory 102.

Continuous data protection journal 108 is operable to retain the write I/O data for access by subsequent read commands and to roll back the data at the primary site. Primary controller 100 further comprises a remote copy controller component 110 adapted either to read the write I/O data from the continuous data protection journal or otherwise to receive the write I/O data—this may be by using a split write technique, for example. Primary controller 100 further comprises a remote copy transmission component 112 operatively coupled to the remote copy controller component 110 to transmit the write I/O data to the secondary 150 of the remote copy pair.

The storage control system described may be operated with a secondary 150 of the remote copy pair, wherein the secondary 150 of the remote copy pair comprises a copy continuous data protection journal 118 to maintain a single copy of each item of write I/O data for use in remote recovery. In such an arrangement, remote copy receive component 114 receives the copy of the write I/O data from the primary and passes it to remote copy control component 116, which writes the copy write I/O data to copy journal 118 and updates directory 120.

The storage control system described may further comprise a compression component to convert the write I/O data to a compressed form and a decompression component to convert the write I/O data in its compressed form to a decompressed form. The compression component may be incorporated into any of the control components at the primary or the secondary, or it may comprise an additional component either in the primary or secondary controller or incorporated into a transmission path of the write I/O data.

The described embodiments may addresses the problem of providing a technological means for maintaining local continuous data protection integrated with a remote copy facility, using commonly-available components and with minimal additional resource use.

With respect to FIG. 4, there is shown in flowchart for one method or logic arrangement of operations. It will be clear to one of ordinary skill in the art that many modifications and variations, including, for example, variations in the sequence of steps, may be made without departing from the scope of the embodiments.

The exemplary method begins as START step 200. The primary controller receives a write command at step 202. At step 204, a test is made to determine whether there is space available in the primary CDP journal 108 (FIG. 3). If space is lacking, or if some threshold amount of available space is not reached, a garbage collector or other means of reclaiming space may be activated at step 206. If space is available, one or more pages are written to the primary CDP journal 108 at step 208, and the primary journal directory 102 is updated accordingly at step 210. At step 212, the one or more pages are sent to the secondary journal 118. At step 214, a test is made to determine whether there is space available in the secondary journal 118. If space is lacking, or if some threshold amount of available space is not reached, a garbage collector or other means of reclaiming space may be activated at step 216. If space is available, the one or more pages are written to the secondary journal 118 at step 218, and the secondary journal directory 120 is updated accordingly at step 220. The exemplary method completes at END step 222.

Certain of the described embodiments may provide a technological means for maintaining local continuous data protection integrated with a remote copy facility, using commonly-available components and with minimal additional resource use.

All or part of the method of the described embodiments may suitably and usefully be embodied in a logic apparatus, or a plurality of logic apparatus, comprising logic elements arranged to perform the operations of the method and that such logic elements may comprise hardware components, firmware components or a combination thereof.

Further, all or part of a logic arrangement may be embodied in a logic apparatus comprising logic elements to perform the described, and that such logic elements may comprise components such as logic gates in, for example a programmable logic array or application-specific integrated circuit. Such a logic arrangement may further be embodied in enabling elements for temporarily or permanently establishing logic structures in such an array or circuit using, for example, a virtual hardware descriptor language, which may be stored and transmitted using fixed or transmittable carrier media.

Yet further, the method and arrangement described above may also suitably be carried out fully or partially in software running on one or more processors (not shown in the figures), and that the software may be provided in the form of one or more computer program elements carried on any suitable data-carrier (also not shown in the figures) such as a magnetic or optical disk or the like. Channels for the transmission of data may likewise comprise storage media of all descriptions as well as signal-carrying media, such as wired or wireless signal-carrying media.

The described embodiments may further be embodied as a computer program product for use with a computer system. Such an implementation may comprise a series of computer-readable instructions either fixed on a tangible medium, such as a computer readable medium, for example, diskette, CD-ROM, ROM, or hard disk, or transmittable to a computer system, using a modem or other interface device, over either a tangible medium, including but not limited to optical or analogue communications lines, or intangibly using wireless techniques, including but not limited to microwave, infrared or other transmission techniques. The series of computer readable instructions embodies all or part of the functionality previously described herein.

Those skilled in the art will appreciate that such computer readable instructions can be written in a number of programming languages for use with many computer architectures or operating systems. Further, such instructions may be stored using any memory technology, present or future, including but not limited to, semiconductor, magnetic, or optical, or transmitted using any communications technology, present or future, including but not limited to optical, infrared, or microwave. It is contemplated that such a computer program product may be distributed as a removable medium with accompanying printed or electronic documentation, for example, shrink-wrapped software, pre-loaded with a computer system, for example, on a system. ROM or fixed disk, or distributed from a server or electronic bulletin board over a network, for example, the Internet or World Wide Web.

In an alternative, the described embodiments may be realized in the form of a computer implemented method of deploying a service comprising steps of deploying computer program code operable to, when deployed into a computer infrastructure and executed thereon, cause said computer system to perform all the method operations.

In a further alternative, the described embodiments may be realized in the form of a data carrier having functional data thereon, said functional data comprising functional computer data structures to, when loaded into a computer system and operated upon thereby, enable said computer system to perform all the steps of the method.

It will be clear to one skilled in the art that many improvements and modifications can be made to the foregoing exemplary embodiments without departing from the scope of the present invention.

Claims

1-16. (canceled)

17. A computer readable storage medium having computer code executed by a primary controller in communication with a secondary controller to perform operations, the operations comprising: retaining modified data for a primary volume as records in a journal, wherein the journal includes records having multiple versions of data;copying records in the journal to the secondary controller;receiving a read request to data, wherein multiple versions of the requested data are included in the journal;determining one of the versions of the requested data to return to the read request;accessing the determined version of the requested data from the journal to return to the read request.

18. The computer readable storage medium of claim 17, wherein the journal comprises a primary journal, and wherein the records in the primary journal are copied to a secondary journal at the secondary controller.

19. The computer readable storage medium of claim 17, wherein the journal includes information linking records in the journal, wherein a plurality of the records provide different versions of data, wherein the read request is with respect to a point-in-time, wherein the operations further comprise: processing the linked records in the journal to determine one record having a timestamp as of the point-in-time of the read request.

20. The computer readable storage medium of claim 19, wherein the operations further comprise: initiating a garbage collection method to remove versions of data from the journal;processing the linked records to determine journal records having previous versions of data to remove; anddeleting the determined journal records.

21. The computer readable storage medium of claim 19, wherein records in the journal each include headers, wherein the linking information comprises at least one pointer in the headers of versions of data to at least one previous version of the data in the record.

22. The computer readable storage medium of claim 21, wherein the journal includes previous versions of data in sequence.

23. The computer readable storage medium of claim 17, wherein each record includes data for one page and a header, wherein the header includes a number of compressed bytes in the page and a timestamp of the data in the page, and wherein the record and data in the record is stored compressed in the journal.

24. The computer readable storage medium of claim 17, wherein the operations further comprise: maintaining a directory having pointers pointing to the records;receiving updated data;including the updated data in a record included the journal;update the pointer in the directory to point to the record included in the journal.

25. The computer readable storage medium of claim 24, wherein the directory and journal comprise a primary directory and primary journal, wherein copying records to the secondary journal causes the secondary controller to copy the records to a secondary journal and update a secondary directory to point to the data written to the secondary journal.

26. A system, comprising: a primary controller having a primary journal;primary volume accessible to the primary controller;a secondary controller in communication with the primary controller;a continuous data protection component executed by the primary controller to perform operations, the operations comprising: retaining modified data for the primary volume as records in the primary journal, wherein the primary journal includes records having multiple versions of data;copying records in the primary journal to the secondary controller;receiving a read request to data, wherein multiple versions of the requested data are included in the primary journal;determining one of the versions of the requested data to return to the read request; andaccessing the determined version of the requested data from the primary journal to return to the read request.

27. The system of claim 26, wherein the operations performed by the continuous data protection component further comprise copying the records in the primary journal are copied to a secondary journal at the secondary controller.

28. The system of claim 26, wherein the primary journal includes information linking records in the journal, wherein a plurality of the records provide different versions of data, wherein the read request is with respect to a point-in-time, wherein the operations performed by the continuous data protection component further comprise: processing the linked records in the primary journal to determine one record having a timestamp as of the point-in-time of the read request.

29. The system of claim 28, wherein records in the primary journal each include headers, wherein the linking information comprises at least one pointer in the headers of versions of data to at least one previous version of the data in the record.

30. The system of claim 26, wherein the operations performed by the continuous data protection component further comprise: maintaining a primary directory having pointers pointing to the records;receiving updated data;including the updated data in a record included the primary journal;update the pointer in the primary directory to point to the record included in the primary journal.

31. The system of claim 30, wherein the secondary controller performs operations, the operations comprising: writing data to a secondary journal; andupdating a secondary directory to point to the data written to the secondary journal.

32. A method, comprising: retaining modified data for a primary volume as records in a journal maintained by a primary controller, wherein the journal includes records having multiple versions of data;copying records in the journal to a secondary controller;receiving, at the primary controller, a read request to data, wherein multiple versions of the requested data are included in the journal;determining one of the versions of the requested data to return to the read request;accessing the determined version of the requested data from the journal to return to the read request.

33. The method of claim 32, the journal comprises a primary journal, and wherein the records in the primary journal are copied to a secondary journal at the secondary controller.

34. The method of claim 32, wherein the journal includes information linking records in the journal, wherein a plurality of the records provide different versions of data, wherein the read request is with respect to a point-in-time, further comprising: processing the linked records in the journal to determine one record having a timestamp as of the point-in-time of the read request.

35. The method of claim 34, wherein records in the journal each include headers, wherein the linking information comprises at least one pointer in the headers of versions of data to at least one previous version of the data in the record.

36. The method of claim 32, further comprising: maintaining a directory having pointers pointing to the records;receiving updated data;including the updated data in a record included the journal;update the pointer in the directory to point to the record included in the journal.

37. The method of claim 36, wherein the directory and journal comprise a primary directory and primary journal, further comprising: writing data, by the secondary controller to the secondary journal; andupdating the secondary directory to point to the data written to the secondary journal.