Search Images Maps Play YouTube News Gmail Drive More »
Sign in
Screen reader users: click this link for accessible mode. Accessible mode has the same essential features but works better with your reader.

Patents

  1. Advanced Patent Search
Publication numberUS20050033930 A1
Publication typeApplication
Application numberUS 10/753,623
Publication dateFeb 10, 2005
Filing dateJan 7, 2004
Priority dateAug 6, 2003
Also published asUS7325112, US20060236051
Publication number10753623, 753623, US 2005/0033930 A1, US 2005/033930 A1, US 20050033930 A1, US 20050033930A1, US 2005033930 A1, US 2005033930A1, US-A1-20050033930, US-A1-2005033930, US2005/0033930A1, US2005/033930A1, US20050033930 A1, US20050033930A1, US2005033930 A1, US2005033930A1
InventorsYumiko Haruma, Koji Sugiyama, Kei Kubo, Yoshihito Yugami, Shinji Morita
Original AssigneeHitachi, Ltd.
Export CitationBiBTeX, EndNote, RefMan
External Links: USPTO, USPTO Assignment, Espacenet
High-speed snapshot method
US 20050033930 A1
Abstract
When a snapshot of data is taken, it is after a write request comes from the host equipment that old data on the source-side magnetic disk to the snapshot-side magnetic disk; therefore, it takes a long time before new data is actually written in the source-side magnetic disk. To overcome this problem, by analyzing data update information, and from a result of analysis, by deciding an area that has high possibility of receiving a request to write data, and having data of that area previously copied to the snapshot-side magnetic disk, it is possible to take a snapshot very quickly.
Images(16)
Previous page
Next page
Claims(16)
1. A method for speeding up a snapshot operation in a computer system wherein a host equipment is connected through a switch to a source-side magnetic disk and a snapshot-side magnetic disk, said method comprising the steps of:
detecting data in an area where data is highly likely to be written next; and
previously copying said data in said snapshot-side magnetic disk.
2. A method for speeding up a snapshot operation in a computer system wherein a host equipment is connected through a switch to a source-side magnetic disk and a snapshot-side magnetic disk, said method further comprising the steps of:
detecting data in an area where data is likely to be written with a possibility higher than a predetermined value; and
copying said data in memory.
3. A snapshot controller for, in response to a request to write data into said source-side magnetic disk from a host equipment, copying a content in an area of said source-side magnetic disk at a point in time to a corresponding area in a snapshot-side magnetic disk, said content being overwritten by new data after said point in time, said snapshot controller comprising:
a magnetic disk read/write executing unit for reading/writing data by a request from said host equipment;
a write-ahead/read-ahead executing unit for writing ahead/reading ahead data in said source-side magnetic disk;
a memory controller for reading and writing necessary information; and
a memory for storing said information.
4. A snapshot controller in claim 3, wherein said memory comprises a magnetic disk information table; a write-ahead information table; a read-ahead information table; and a data update information table.
5. A snapshot controller in claim 4, wherein said magnetic disk information table includes host equipment information showing which host equipment sent a read/write request; magnetic disk pair information showing a pair of magnetic disks where a snapshot operation is performed; area information showing how many sectors of data are read ahead/written ahead in which areas; and read/write information showing whether data is read or written.
6. A snapshot controller in claim 4, wherein said write-ahead information table includes source-side magnetic disk number information showing data of which magnetic disk is written ahead; and data write-ahead area information showing how many sectors of data are of which areas are written ahead.
7. A snapshot controller in claim 4, wherein said read-ahead information table includes source-side magnetic disk number information showing data of which magnetic disks is read ahead; data read-ahead area information showing how many sectors of data of which areas are read ahead; and read-ahead data showing data read ahead.
8. A method for controller in claim 4, wherein said data update information table includes host equipment number information showing from which host equipment a write request came; source-side magnetic disk number information showing the magnetic disks new data was written; and data write area information showing the areas where new data was written and how many sectors of data were written.
9. A snapshot controller in claim 3, wherein said magnetic disk read/write executing unit comprises means for obtaining information about the areas where data was written ahead in the snapshot-side magnetic disk; and means for deciding the write-ahead/read-ahead state to see if the area the host equipment is making a data write request coincides with the area where data was written ahead or read ahead.
10. In the snapshot controller in claim 4, wherein said write-ahead/read-ahead executing unit comprises means for deciding if read-ahead data is stored in said read-ahead information table; means for, when read-ahead data is stored in the read-ahead information table, copying read-ahead data to the snapshot-side magnetic disk; and means for recording the area data of which was copied as the write-ahead area in said write-ahead information table.
11. A snapshot controller in claim 4, wherein said write-ahead/read-ahead executing unit comprises means for obtaining log information about writing data to the source-side magnetic disk, from said data update information table; means for analyzing obtained information and deciding a write-ahead area and data capacity for writing ahead; and means for writing ahead data in a decided write-ahead area to said snapshot-side magnetic disk.
12. A snapshot controller in claim 4, wherein said write-ahead/read-ahead executing unit comprises means for obtaining log information about writing data to said source-side magnetic disk from said data update information table; means for analyzing obtained information and deciding a read-ahead area and data capacity for reading ahead; and means for storing data in a decided read-ahead area in said read-ahead information table.
13. A host equipment for, by issuing a request to write data into said source-side magnetic disk, copying a content in an area of said source-side magnetic disk at a point in time to a corresponding area in a snapshot-side magnetic disk, said content being overwritten by new data after said point in time, said host equipment comprising:
magnetic disk read/write executing unit for reading/writing data according to said data read/write request;
a memory controller for reading and writing necessary information from a memory; and
a memory for storing necessary information.
14. A switch for, in response to host equipment's request to write data to a source-side magnetic disk, copying a content in an area of said source-side magnetic disk at a point in time to a corresponding area in a snapshot-side magnetic disk, said content being overwritten by new data after said point in time, said switch comprising:
a magnetic disk read/write executing unit for reading/writing data according to the request from said host equipment;
a write-ahead/read-ahead executing unit for writing/reading data according to the request of said host equipment;
a memory controller for reading and writing necessary information from a memory; and
a memory for storing necessary information.
15. A storage system for, in response to host equipment's request to write data to a source-side magnetic disk, copying a content in an area of said source-side magnetic disk at a point in time to a corresponding area in a snapshot-side magnetic disk, said content being overwritten by new data after said point in time, said storage system comprising:
a write-ahead/read-ahead executing unit for writing/reading data according to the request of said host equipment;
a memory controller for reading and writing necessary information from a memory; and
a memory for storing necessary information.
16. A snapshot controller comprises:
means for discriminating between an area from which data was copied by a data write request from host equipment and an area from which data was copied by a write-ahead operation by using two kinds of flags when old data in the source-side magnetic disk was copied to the snapshot-side magnetic disk.
Description
    BACKGROUND OF THE INVENTION
  • [0001]
    The present invention relates to a high-speed snapshot function of copying a content in an area of a source-side magnetic disk at a point in time to a corresponding area in a snapshot-side magnetic disk, the content being overwritten by new data after the above-mentioned point in time, and also to host equipment, a switch, and a storage system as means for realizing the high-speed snapshot.
  • [0002]
    In the prior art of snapshot technology, when a request to write data to the source-side magnetic disk comes from the host equipment, data present in the area is copied to the snapshot-side magnetic disk. After copying data to the snapshot-side is finished, new data is written in the source-side magnetic disk. When the host equipment again issues a request to write data to the source-side magnetic disk, just as mentioned above, data in the specified area is copied to the snapshot-side magnetic disk, and then new data is written to the source-side magnetic disk. This technology is disclosed in U.S. Pat. No. 5,649,152.
  • [0003]
    The snapshot copy, such as mentioned above, can be controlled by the host equipment or the storage system.
  • [0004]
    As a method for managing snapshot, there is a technology in which in a case a snapshot is not taken, when data is written from host equipment to an external storage device, data is written in a plurality of storage media to thereby provide a couple of storage areas in which data is stored, and when a snapshot is taken, one of the paired storage areas is used as an area for ordinary access, and the other storage area is used for snapshot. This technology is disclosed in JP-A-2001-306407.
  • SUMMARY OF THE INVENTION
  • [0005]
    In the prior art, it is after a write request comes from the host equipment that old data on the source-side magnetic disk is copied to the snapshot-side magnetic disk; therefore, it takes time before new data is actually written in the source-side magnetic disk. The method disclosed in JP-A-2001-306407 is a snapshot managing method in the host equipment, for which reason it is impossible to manage snapshots by the switch or the storage system.
  • [0006]
    To solve this problem, the present invention has as its object to provide a high-speed snapshot method, host equipment, a switch, and a storage system.
  • [0007]
    A write-ahead/read-ahead function for a high-speed snapshot is added to the host equipment for controlling snapshot, the switch and the storage system. More specifically, the present invention provides means for analyzing a data writing history, means for automatically deciding a write-ahead area and a read-ahead area from a result of analysis, means for copying data in a decided write-ahead area to the snapshot-side magnetic disk before new data is written in that write area, and means for previously storing data in a decided read-ahead area into a memory.
  • BRIEF DESCRIPTION OF THE DRAWINGS
  • [0008]
    FIG. 1 is a diagram of a structure for realizing speeding up of a snapshot;
  • [0009]
    FIG. 2 is a magnetic disk information table;
  • [0010]
    FIG. 3 is a write-ahead information table;
  • [0011]
    FIG. 4 is a read-ahead information table;
  • [0012]
    FIG. 5 is a data update information table;
  • [0013]
    FIG. 6 shows a schematic procedure of speeding up of a snapshot;
  • [0014]
    FIG. 7 is a flowchart showing a process of deciding the states of write-ahead/read-ahead areas and writing data;
  • [0015]
    FIG. 8 is a flowchart showing a write-ahead/read-ahead process;
  • [0016]
    FIG. 9 is a flowchart (Part 1) showing a process of deciding write-ahead/read-ahead areas;
  • [0017]
    FIG. 10 is a flowchart (Part 2) showing a process of deciding a write-ahead/read-ahead area;
  • [0018]
    FIG. 11 is a flowchart (Part 3) showing a process of deciding a write-ahead/read-ahead area;
  • [0019]
    FIG. 12 is a flowchart (Part 4) showing a process of deciding a write-ahead/read-ahead area;
  • [0020]
    FIG. 13 shows an example of writing data by sequential access;
  • [0021]
    FIG. 14 shows an example of writing data by random access;
  • [0022]
    FIG. 15 shows an example of writing data by access frequency analysis;
  • [0023]
    FIG. 16 shows an example of executing a write-ahead/read-ahead;
  • [0024]
    FIG. 17 shows an example of a data update information table;
  • [0025]
    FIG. 18 shows an example of a data update information table; and
  • [0026]
    FIG. 19 shows an embodied example of returning the copy.
  • DESCRIPTION OF THE EMBODIMENT
  • [0027]
    A preferred embodiment will be described with reference to FIGS. 1˜18.
  • [0028]
    In the following description, a write-ahead indicates a function by which to have previously placed in a snapshot-side magnetic disk, data in an area where data is highly likely to be written (hereafter referred to as a write-ahead area). A read-ahead indicates a function by which to place in memory, data in an area where data is likely to be written with likelihood second highest next to the above-mentioned write-ahead area (hereafter referred to a read-ahead area).
  • [0029]
    In this embodiment, snapshot is controlled by a switch, and it is assumed that write-ahead/read-ahead is carried out after new data has been written in a source-side magnetic disk in response to a write request from host equipment, such as a host computer.
  • [0030]
    Description will start with the structure of the present invention with reference to FIG. 1.
  • [0031]
    The host equipment 110 is connected to a source-side magnetic disk 150 and a snapshot-side magnetic disk 160 by a switch 120. The host equipment 110 can access data in the source-side magnetic disk 150 and the snapshot-side magnetic disk through the switch 120. The switch 120 is a snapshot controller that embodies the present invention and, by a request from the host equipment, reads data from the source-side magnetic disk 150 and places in the snapshot-side magnetic disk 160, a copy of old data in the area to write data in. The switch 120 comprises a magnetic disk read/write executing unit 121 for reading/writing data by a request from the host equipment, a write-ahead/read-ahead executing unit 124 for writing ahead or reading ahead data in the source-side magnetic disk, a memory controller 122 for reading or writing necessary information to and from memory, and a memory 123 for storing necessary information. The memory 123 comprises a magnetic disk information table 125, a write-ahead information table 126, a read-ahead information table 127, and a data update information table 128.
  • [0032]
    FIG. 2 is the magnetic disk information table 125. This table includes host equipment information 201 for showing from which host equipment a read/write request came, magnetic disk pair information 202 for snapshot operation, area information 203 for showing how many sectors of data are written or read in which area, and read-ahead/write-ahead information 204 for showing whether to read or write.
  • [0033]
    FIG. 3 is a write-ahead information table 126. This table includes source-side magnetic disk numbers 301 showing data of which magnetic disk is written ahead, and data write-ahead area information 302 for showing how many sectors of data of which area are written ahead.
  • [0034]
    FIG. 4 is a read-ahead information table 127. This table includes source-side magnetic disk numbers 401 for showing data of which magnetic disk is read ahead, data read-ahead area information 402 for showing how many sectors of data of which magnetic disk are read ahead, and read-ahead data 403 for showing data read ahead.
  • [0035]
    FIG. 5 is a data update information table 128. This table includes source-side magnetic disk numbers 502 for showing which host equipment a write request came from, source-side magnetic disk numbers 502 for showing a magnetic disk where new data was written, and data write area information 503 for showing an area where new data was written and the number of sectors of data written. By analyzing the data update log recorded in this table, it is possible to decide a write-ahead and a read-ahead area.
  • [0036]
    Referring to FIG. 6, description will be made of a schematic procedure of a method of speeding up snapshot according to the present invention.
  • [0037]
    The host equipment is connected through the switch 120 to the source-side magnetic disk 150 and the snapshot-side magnetic disk 160. On receiving a request to write data in an area A of the source-side magnetic disk 150 from the host equipment 110 (S601), the switch 120 copies old data of the source-side magnetic disk 150 (S602) and writes new data in the area A of the source-side magnetic disk 150 (S603). Then, the switch 120 performs a process for high-speed snapshot as follows. The switch 120 analyzes the areas where data were written by that point in time and decides a write-ahead and a read-ahead area. In this example, it is assumed that the write-ahead area is decided as an area B located continuous to the area A. It is also assumed that the read-ahead area is decided as an area C continuous to the area B.
  • [0038]
    The switch 120 previously copies old data of the decided write-ahead area B to the snapshot-side magnetic disk (S604). Old data in the read-ahead area C is stored in a memory 123 in the switch 120 (S605).
  • [0039]
    When a request comes again from the host equipment 110 to write data in the area B of the source-side magnetic disk 150 (S606), because the data in the area B was already copied to the snapshot-side magnetic disk 160 in Step S604, the switch 120 can immediately write new data to the area B of the source-side magnetic disk 150 (S607).
  • [0040]
    Further, when a request comes again from the host equipment 110 to write data in the area C of the source-side magnetic disk 150 (S608), because the data in the area C has been stored in the memory 123 in the switch 120 in S605, the switch 120, without accessing the area C of the source-side magnetic disk 150, can quickly copy data in the memory 123 to the snapshot-side magnetic disk 160 (S609) and write new data to the area C of the source-side magnetic disk 150 (S610).
  • [0041]
    Various operations will be described in detail with reference to FIGS. 1, 7˜12.
  • [0042]
    FIG. 7 is a flowchart showing a process of deciding the states of write-ahead/read-ahead areas and writing data in the magnetic disk read/write executing unit 121.
  • [0043]
    When a data write request comes from the host equipment 110, the switch 120 obtains from the write-ahead-information table 126 information about the areas data of which has been written ahead to the snapshot-side magnetic disk 160 (S701). A decision is made if the area the host equipment is making a request to write data in coincides with the area data on which has already been copied to the snapshot-side magnetic disk (S702). If the result of the decision shows that the area data is to be written in coincides with the area from which data was copied, new data is written to the source-side magnetic disk 150, by which the process is finished (S707).
  • [0044]
    As a result of the decision, if there is not coincidence about the areas, information about the area data on which was already stored in the memory by a read-ahead operation is obtained from the read-ahead information table 127 (S703). And, a decision is made if the area where the host equipment is making a request to write data coincides with the area from which data was already copied by a read-ahead operation (S704). If the result of decision shows the area where data is to be written coincides with the area from which data was copied by a read-ahead operation, read-ahead data previously recorded in the read-ahead information table 127 in the memory 123 is copied to the snapshot-side magnetic disk 160 (S706), and new data is written to the source-side magnetic disk 150 (S707), by which the process is finished.
  • [0045]
    As a result of decision, if the area where data is to be written does not coincide with either the write-ahead area or the read-ahead area, as usual the data on the source-side magnetic disk 150 is copied to the snapshot-side magnetic disk 160 (S705), then new data is written to the source-side magnetic disk 150 (S707), by which the process is finished.
  • [0046]
    FIG. 8 is a flowchart showing the write-ahead/read-ahead process in the write-ahead/read-ahead executing unit 124 in FIG. 1. As shown in FIG. 7, after new data is written in the source-side magnetic disk 150, the write-ahead area and the read-ahead area are recorded in the respective tables.
  • [0047]
    In this embodiment, description will be made of a case where after new data is written into the source-side magnetic disk 150 by a write request from the host equipment 110 (S801), a write-ahead and a read-ahead are carried out. From the magnetic disk information table 125, information is obtained about the magnetic disks that form the above-mentioned pair of magnetic disks (S802). After this, a write-ahead area and a read-ahead area are decided (S803). From the read-ahead information table 127, source-side magnetic disk numbers, read-ahead area information and read-ahead data are obtained (S804). By using information obtained at S804, a decision is made if data in the write-ahead area decided in step S803 has been already stored in the memory 123 (S805).
  • [0048]
    As a result of decision, if data in the write-ahead area decided in step S803 has been already stored in the memory 123 by a previous data writing operation, the data in the memory is copied to the snapshot-side magnetic disk (S806), and the area from which data was copied is recorded as a write-ahead area in the write-ahead information table 126 (S808). As a result of decision, if data in the write-ahead area decided in step S803 is not stored in the memory 123, the data in the write-ahead area is copied to the snapshot-side magnetic disk 160 (S807), and the area from which data was written ahead is recorded in the write-ahead information table (S808). Subsequently, information about the read-ahead area decided in step S803, including source-side magnetic disk numbers, data read-ahead area information, and data in read-ahead area, is recorded in the read-ahead information table 127 (S809).
  • [0049]
    Referring to FIGS. 9˜16, detailed description will be made of a method for deciding a write-ahead area and a read-ahead area.
  • [0050]
    FIGS. 9˜12 are flowcharts of deciding a write-ahead area and a read-ahead area (S803) in FIG. 8.
  • [0051]
    To decide a write-ahead area and a read-ahead area, data-write log information to the source-side magnetic disk 150 is obtained from data-update information table 128 (S901). To be more specific, the log information includes host equipment numbers, source-side magnetic disk numbers, and data write areas. The host equipment numbers show from which hosts came requests to write. The source-side magnetic disk numbers show the numbers of source-side magnetic disks where data was written. The data write area information shows the location where data writing began by LBA (Logical Block Addressing), and data-writing capacity in sector units. By using LBA information, the magnetic disk access method specific to the application in the host equipment is analyzed (S902).
  • [0052]
    If this analysis shows that data was written by sequential access (S903), the area continuously next to the latest area where data was written is decided as a write-ahead area (S907). And, the area continuously next to the write-ahead area is decided as a read-ahead area (S908). If random access was used for writing data (S904) in other areas, an interval (a difference in area number) between an area where data was written and the next area where data was written is calculated (S909). By using the result of calculation, the area of an area number counted starting from the number of the latest data-written area (this area number is the area number of the latest data-written area+the difference calculated at S909) is decided as a write-ahead area (S910). On the other hand, the area of “the area number of the above-mentioned write-ahead area+the difference calculated at S909” is decided as a read-ahead area (S911).
  • [0053]
    When analyzing the method for writing data by access frequency analysis (S905), the areas where data was written with high frequency are analyzed by using information obtained in S901 (S912), the method by which data was written in the areas where the writing frequency is higher than a predetermined value is analyzed (S913); for example, in a case where data is to be written by sequential access (S914), steps S907 and S908 are carried out, and in a case where data is to be written by random access (S915), steps S909 to S911 are carried out. After this, by using obtained data-writing capacity information, a decision is made how many sectors of data are written ahead or read ahead (S906). For example, if the write capacity information in the data write area information in the data update information table 128 has the area divided in units of 20 sectors, a decision is derived that it is highly likely that the average value of data to be written next is 20 sectors, and the amount of data to write ahead or read ahead is set as 20 sectors.
  • [0054]
    FIG. 13 shows an example of data writing by sequential access.
  • [0055]
    The first data writing (S1001) is performed in the area 1001, the second data writing (S1002) in the area 1002, the third data writing (S1003) in the area 1003, and the fourth data writing (S1004) in the area 1004. Because the latest area where data was written was the area 1004, the next write-ahead area is decided as the area 1005 continuously next to the area 1004. The read-ahead area is decided as the area 1006 continuously next to the write-ahead area.
  • [0056]
    FIG. 14 shows an example of data writing by random access.
  • [0057]
    The first data writing (S1101) is performed in the area 1101, the second data writing (S1102) in the area 1105, the third data writing (S1103) in the area 1109, and the fourth data writing (S1104) in the area 1102. Because the latest area where data was written was the area 1102, the write-ahead area is decided as the area 1106, the fourth area from the area 1102, and the read-ahead area is decided as the area 1110, the fourth area from the write-ahead area.
  • [0058]
    FIG. 15 shows an example of data writing by access frequency analysis. The first data writing (S1201) is performed in the area 1201, the second data writing (S1202) in the area 1202, the third data writing (S1203) in the area 1203, the fourth data writing (S1204) in the area 1204, the fifth data writing (S1205) in the area 1201, and the sixth data writing (S1206) in the area 1202. The frequently-accessed areas are the areas 1201˜1204, and the latest area where data was written is the area 1202. In the areas 1201˜1204, the data access method is sequential access. Therefore, the write-ahead area is the area 1203 continuously next to the area 1202, and the read-ahead area is the area 1204 continuously next to the write-ahead area.
  • [0059]
    FIG. 16 shows an example of execution of a write-ahead and a read ahead. When a request comes from the host equipment 110 to write data to the sectors 0 and 1 of the source-side magnetic disk (S1301), data stored in this area is copied to the sectors 0 and 1 of the snapshot-side magnetic disk, and then new data is written into the source-side magnetic disk (S1303). After this, data on the sectors 2 and 3 of the source-side magnetic disk as the write-ahead area is copied to the sectors 2 and 3 of the snapshot-side magnetic disk (S1304). And, data on the sectors 4 and 5 of the source-side magnetic disk as the read-ahead area is copied to the memory 123 in the switch 120 (S1305).
  • [0060]
    For deciding a write-ahead operation and a read-ahead operation, methods other than shown above may be used, and they will be described with reference to FIGS. 17 and 18.
  • [0061]
    In FIG. 17, data written in the data write capacity in the data update information table 128 is stored in byte units. For example, if the write-in capacity information has data written in units of 20 bytes, the data capacity to write ahead and read ahead is decided as 20 bytes, the average value of those byte-units.
  • [0062]
    FIG. 18 shows a case where the write-ahead and read-ahead areas are decided with the data capacity for write-ahead and read-ahead operations fixed. By using LBA information obtained, the write-ahead and read-ahead areas are decided as shown in FIGS. 9˜12 and data corresponding to a predetermined data capacity are written ahead and read ahead.
  • [0063]
    In this embodiment, the snapshot function is controlled by the switch 120 but may be controlled by the host equipment or the storage system.
  • [0064]
    Description will now be made of the snapshot control procedure with the host equipment 110. After deciding the write-ahead and read-ahead areas, the host equipment 110, through the switch 120, reads data in the write-ahead and read-ahead areas of the source-side magnetic disk 150, and copies data on the write-ahead area to the snapshot-side magnetic disk 160, and stores data read ahead in the memory of the host equipment 110. After this, new data is written to the source-side magnetic disk 150.
  • [0065]
    Description will then be made of the snapshot control procedure with the storage system. The storage system receives information about the location to write data in along with data to write, from the host equipment 110 through the switch 120. Subsequently, after the write-ahead and read-ahead areas are decided, data in the write-ahead area is copied to the snapshot-side magnetic disk 160, and data to read ahead is stored in the cache in the storage system. New data received from the host equipment 110 is written to the source-side magnetic disk 150.
  • [0066]
    When the snapshot function is controlled by the storage system, but if the storage system does not have a cache, if data to read ahead is stored in the memory of the switch 120 or in the memory of the host equipment 110, the read-ahead process can be carried out.
  • [0067]
    In this embodiment, the write-ahead and read-ahead processes are executed after new data has been written in the source-side magnetic disk 150; however, the write-ahead and read-ahead processes may be executed with various timing. For example, while waiting for a request to input/output data from the host equipment 110, when a request comes from the host equipment 110 to write ahead/read ahead, the write-ahead and read-ahead processes can be carried out.
  • [0068]
    In this embodiment, description has been made referring to a case where magnetic disks are used as storage devices, but this embodiment may be applied to magnetic tape, optical disks and so on.
  • [0069]
    The present invention may be applied to other embodiments, such as a case where the source-side magnetic disk is brought back to the state before data was copied (hereafter referred to as copy return).
  • [0070]
    FIG. 19 shows an example of copy return.
  • [0071]
    A bit map 1630 uses two kinds of flag “1” or “2” for every area of the snapshot-side magnetic disk. Description will be made of a case where the copy can be returned easily by using those flags.
  • [0072]
    When a data write request comes from the host equipment 110, old data on the source-side magnetic disk 1610 is copied to the snapshot-side magnetic disk 1620 (S1601), and when new data is written in the source-side magnetic disk 1610, the “1” flag is set in the bit map 1630. By a write-ahead process (S1602), data in the write-ahead area of the source-side magnetic disk 1610 is copied to the snapshot-side magnetic disk 1620, the “2” flap is set in the bit map 1630.
  • [0073]
    After this, if it is necessary to execute copy return (S1603), only data in the areas marked by the “1” flag is returned to the source-side magnetic disk 1640. Because data in the areas marked by the “2” flag is data which was copied by “a write ahead” to the snapshot-side magnetic disk 162, new data has not been written to the source-side magnetic disk 1610, and data in the area marked by the “2” flag need not be returned to the source-side magnetic disk 1640.
  • [0074]
    By the method described above, it is easy to perform a process to bring the source-side magnetic disk 1610 back to the state 1640 before data was copied from there.
  • [0075]
    The program for executing a method for speeding up the snapshot operation in the present invention described above can be stored in a storage medium readable by a computer, or transferred to another computer through communication means, and the program can be read into memory and executed.
  • [0076]
    By analyzing data update information and deciding from the result of analysis an area where chances are high for a request to come to write data to the source-side magnetic disk, and executing a write-ahead process of writing in advance data in that area, it is possible to write data to the source-side magnetic disk on receiving a write request from the host equipment. Furthermore, by analyzing data update information and deciding by result of analysis an area where it is considerably likely for a request to come to write into the source-side magnetic disk, and by executing a read-ahead process of previously reading data in that area and storing it in memory, it is possible to copy data in memory into the snapshot-side magnetic disk quickly when a write request comes from the host equipment. By performing write-ahead and read-ahead processes in combination, the snapshot operation can be accelerated. Moreover, by using two kinds of flags in the bit map, the source-side magnetic disk can be reverted to the state before data was copied.
  • [0077]
    It should be further understood by those skilled in the art that although the foregoing description has been made on embodiments of the invention, the invention is not limited thereto and various changes and modifications may be made without departing from the spirit of the invention and the scope of the appended claims.
Patent Citations
Cited PatentFiling datePublication dateApplicantTitle
US5649152 *Oct 13, 1994Jul 15, 1997Vinca CorporationMethod and system for providing a static snapshot of data stored on a mass storage system
US6678809 *Apr 13, 2001Jan 13, 2004Lsi Logic CorporationWrite-ahead log in directory management for concurrent I/O access for block storage
US6807605 *Oct 3, 2002Oct 19, 2004Hewlett-Packard Development Company, L.P.Managing a data storage array, a data storage system, and a raid controller
US6907505 *Jul 31, 2002Jun 14, 2005Hewlett-Packard Development Company, L.P.Immediately available, statically allocated, full-logical-unit copy with a transient, snapshot-copy-like intermediate stage
US20040181642 *Feb 3, 2004Sep 16, 2004Haruaki WatanabeStorage system and snapshot management method thereof
US20050033800 *Jun 25, 2004Feb 10, 2005Srinivas KavuriHierarchical system and method for performing storage operations in a computer network
Referenced by
Citing PatentFiling datePublication dateApplicantTitle
US7634507Aug 30, 2006Dec 15, 2009Inmage Systems, Inc.Ensuring data persistence and consistency in enterprise storage backup systems
US7676502Mar 9, 2010Inmage Systems, Inc.Recovery point data view shift through a direction-agnostic roll algorithm
US7698401Jun 1, 2004Apr 13, 2010Inmage Systems, IncSecondary data storage and recovery system
US7979656Aug 6, 2008Jul 12, 2011Inmage Systems, Inc.Minimizing configuration changes in a fabric-based data protection solution
US8028194Jul 25, 2008Sep 27, 2011Inmage Systems, IncSequencing technique to account for a clock error in a backup system
US8055745Sep 16, 2005Nov 8, 2011Inmage Systems, Inc.Methods and apparatus for accessing data from a primary data storage system for secondary storage
US8069227Dec 26, 2008Nov 29, 2011Inmage Systems, Inc.Configuring hosts of a secondary data storage and recovery system
US8224786Dec 26, 2008Jul 17, 2012Inmage Systems, Inc.Acquisition and write validation of data of a networked host node to perform secondary storage
US8321377Apr 17, 2006Nov 27, 2012Microsoft CorporationCreating host-level application-consistent backups of virtual machines
US8527470Dec 26, 2008Sep 3, 2013Rajeev AtluriRecovery point data view formation with generation of a recovery view and a coalesce policy
US8527721Dec 26, 2008Sep 3, 2013Rajeev AtluriGenerating a recovery snapshot and creating a virtual view of the recovery snapshot
US8554727May 19, 2006Oct 8, 2013Inmage Systems, Inc.Method and system of tiered quiescing
US8601225Dec 26, 2008Dec 3, 2013Inmage Systems, Inc.Time ordered view of backup data on behalf of a host
US8683144Dec 26, 2008Mar 25, 2014Inmage Systems, Inc.Causation of a data read against a first storage system to optionally store a data write to preserve the version to allow viewing and recovery
US8838528Dec 26, 2008Sep 16, 2014Inmage Systems, Inc.Coalescing and capturing data between events prior to and after a temporal window
US8868858May 19, 2006Oct 21, 2014Inmage Systems, Inc.Method and apparatus of continuous data backup and access using virtual machines
US8949395Jul 24, 2009Feb 3, 2015Inmage Systems, Inc.Systems and methods of event driven recovery management
US9098455Sep 25, 2014Aug 4, 2015Inmage Systems, Inc.Systems and methods of event driven recovery management
US9209989Dec 26, 2008Dec 8, 2015Inmage Systems, Inc.Causation of a data read operation against a first storage system by a server associated with a second storage system according to a host generated instruction
US20060010227 *Sep 16, 2005Jan 12, 2006Rajeev AtluriMethods and apparatus for accessing data from a primary data storage system for secondary storage
US20060031468 *Jun 1, 2004Feb 9, 2006Rajeev AtluriSecondary data storage and recovery system
US20070271304 *May 19, 2006Nov 22, 2007Inmage Systems, Inc.Method and system of tiered quiescing
US20070271428 *May 19, 2006Nov 22, 2007Inmage Systems, Inc.Method and apparatus of continuous data backup and access using virtual machines
US20070282921 *May 22, 2006Dec 6, 2007Inmage Systems, Inc.Recovery point data view shift through a direction-agnostic roll algorithm
US20080059542 *Aug 30, 2006Mar 6, 2008Inmage Systems, Inc.Ensuring data persistence and consistency in enterprise storage backup systems
US20100023797 *Jan 28, 2010Rajeev AtluriSequencing technique to account for a clock error in a backup system
US20100169281 *Dec 26, 2008Jul 1, 2010Rajeev AtluriCoalescing and capturing data between events prior to and after a temporal window
US20100169282 *Dec 26, 2008Jul 1, 2010Rajeev AtluriAcquisition and write validation of data of a networked host node to perform secondary storage
US20100169452 *Dec 26, 2008Jul 1, 2010Rajeev AtluriCausation of a data read operation against a first storage system by a server associated with a second storage system according to a host generated instruction
US20100169466 *Dec 26, 2008Jul 1, 2010Rajeev AtluriConfiguring hosts of a secondary data storage and recovery system
US20100169587 *Dec 26, 2008Jul 1, 2010Rajeev AtluriCausation of a data read against a first storage system to optionally store a data write to preserve the version to allow viewing and recovery
US20100169591 *Dec 26, 2008Jul 1, 2010Rajeev AtluriTime ordered view of backup data on behalf of a host
US20100169592 *Dec 26, 2008Jul 1, 2010Rajeev AtluriGenerating a recovery snapshot and creating a virtual view of the recovery snapshot
WO2007130192A1 *Feb 15, 2007Nov 15, 2007Microsoft CorpCreating host-level application-consistent backups of virtual machines
WO2015066698A1 *Nov 4, 2014May 7, 2015Falconstor, Inc.Snapshots using copy on predicted write
Classifications
U.S. Classification711/162
International ClassificationG06F12/16, G06F3/06, G06F11/14
Cooperative ClassificationG06F2201/84, G06F3/0653, G06F3/065, G06F3/067, G06F11/1466, G06F3/0611
European ClassificationG06F3/06A4M, G06F3/06A6D, G06F3/06A2P2, G06F3/06A4H4
Legal Events
DateCodeEventDescription
Jun 21, 2004ASAssignment
Owner name: HITACHI, LLTD., JAPAN
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:HARUMA, YUMIKO;REEL/FRAME:015493/0347
Effective date: 20040223