JP4788492B2 - Storage device capacity expansion method, program, and storage device - Google Patents

Description

  The present invention relates to a storage apparatus capacity expansion method, program, and storage apparatus, and in particular, a storage apparatus in which a plurality of logical volumes are set on a RAID (Redundant Arrays of Inexpensive Disks) group composed of a plurality of physical disks. The present invention relates to a capacity expansion method, a program, and a storage apparatus for a storage apparatus capable of expanding the active capacity (Logical Device Expansion, LDE) of a logical volume by adding a physical disk.

Conventionally, there has been disclosed a method of expanding the storage capacity by physically adding a disk even in an active state. (See Patent Documents 1 and 2)
Japanese Patent Laid-Open No. 8-212018 FIG. 10 shows a state of data blocks and parity blocks (conventional example 1) on a (physical) disk before capacity expansion described in Japanese Patent Laid-Open No. 2000-511322. These show the state of the rearranged data block and parity block on the (physical) disk after capacity expansion in Conventional Example 1.

  FIG. 12 shows the state of data blocks and parity blocks on a (physical) disk on a disk (disks 0, 1, and 2) before capacity expansion described in Patent Document 2 (conventional example 2). 13 shows the state of data blocks and parity blocks on the (physical) disk after capacity expansion in Conventional Example 2, and FIG. 14 shows the constraint conditions in Conventional Example 2.

  In Conventional Example 2, the storage capacity expansion of the level 5 RAID array is: A) a new data block (D10) without parity blocks present on the added set of disks (disks 3 and 4) in the array row. To D19) and B) as shown in FIG. 14, initializing new data blocks such that the exclusive OR of their contents in each row of the array is zero. .

  In a RAID storage apparatus, generally, on the physical disks P # 0, P # 1, P # 2, and P # 3, as shown on the left side (RAID group before LDE) of the capacity expansion example (equal expansion) shown in FIG. A plurality of logical volumes, for example, a logical volume L # 0 with a capacity of m0 GB (gigabytes), a logical volume L # 1 with a capacity of m1 GB, and a logical volume L # 2 with a capacity of m2 GB are set and operated.

  In such an operation, the capacity can be expanded by adding the physical disk # P4. The example on the right side of FIG. 15 (RAID group after LDE) shows an example of expanding the capacity of each logical volume equally.

  On the other hand, FIG. 16 shows a capacity expansion example (logical volume addition) in which a new logical volume L # 3 is added without changing the capacity of each logical volume, and FIG. 17 shows only the last logical volume (FIG. 17). In the example, only the logical volume L # 2) is shown as an example of capacity expansion (final logical volume capacity expansion).

  In the example of the equal expansion shown in FIG. 15, the physical disk P is added to each logical volume by adding to the logical block address (LBA) on the physical disks P # 0 to P # 3 allocated to each logical volume before capacity expansion. If the LBA of # 4 is evenly distributed and rearranged, data rearrangement occurs in the area of each logical volume, so even if each logical volume data is stored, it interferes with the data of other logical volumes. Never do.

  In the example of adding a logical volume shown in FIG. 16, the processing of the logical volume L # 0 is the same as the processing of the example shown in FIG. In the processing of the logical volume L # 1, the head LBA is arranged to overwrite the area of the logical volume L # 0 before capacity expansion, but the data of the logical volume L # 0 has already been rearranged. There is no data destruction. The same applies to the processing of the logical volume L # 2.

  The example of the last logical volume capacity expansion shown in FIG. 17 is the same as the example shown in FIG. 16 in the middle, and all the capacity expanded by adding the physical disk P # 4 is added to the last logical volume L # 2. Therefore, it will not cause data corruption.

  However, in actual operation, as shown in FIG. 18, the capacity of the logical volume arranged on the front side of the physical disk, for example, the capacity of the first logical volume is expanded, or the capacity of a plurality of logical volumes is increased as shown in FIG. There are various requirements such as expansion. Even in such a case, it is necessary to perform capacity expansion by rearranging the data stored in each logical volume without destroying it.

  In the case of the example shown in FIG. 18, when data is simply rearranged in the LBA order according to the capacity designation after capacity expansion, the data of the logical volume L # 1 is stored in the area of the logical volume L # 2 before the rearrangement as shown in FIG. Is overwritten, the data of the logical volume L # 2 is destroyed.

  In the method described in Patent Document 1, when data blocks and parity blocks on a physical disk are rearranged, new parity is calculated and written to the extended array. However, the data is set on the physical disk. The logical volume is not considered.

  In the method described in Patent Document 2, for example, A) new data blocks (D10 to D19) without parity blocks existing on the added set of disks (disks 3 and 4) in the rows of the array. ) Eliminates the need to rewrite the original data blocks and parity blocks on the first set of disks, but does not take into account logical volumes set up on physical disks .

  In the present invention, when capacity expansion is performed by adding a physical disk to a storage apparatus in which a plurality of logical volumes are set on a RAID group configured with a plurality of physical disks, the stored data is not destroyed. It is an object of the present invention to provide a method, a program, and a storage device that perform capacity expansion.

  The first invention adds a physical disk by specifying the capacity of each logical volume after capacity expansion to a storage device in which a plurality of logical volumes are set on a RAID group composed of a plurality of physical disks. The capacity of each logical volume after capacity expansion is received, and the beginning and end LBA of each logical volume before and after capacity expansion are obtained in order from the logical volume set to the top LBA. The logical volume that is registered in the capacity expansion management table and the first LBA and the last LBA after the capacity expansion are not moved later is registered in the capacity expansion management table with the order as “normal order”, and the first LBA does not move after the capacity expansion, The logical volume to which the last LBA moves later is the logical volume at a position where the LBA does not change before and after capacity expansion. Divide and register the first half order as “normal order” and the second half order as “reverse order” in the capacity expansion management table. After the capacity expansion, the logical volume to which the first LBA and the last LBA move later is designated as “reverse order” For logical volumes that are registered in the capacity expansion management table, the first LBA moves backward after capacity expansion, and the last LBA does not move later, the logical volume is divided at a position where the LBA does not change before and after capacity expansion, and the first half is “reverse order”. If the second half is registered in the capacity expansion management table as “normal order” and a physical disk is added, the order of the first logical volume registered in the capacity expansion management table is read. If the order is “normal order”, If the logical volume moves from the first LBA before capacity expansion to the LBA after capacity expansion in order, and the order is “reverse order”, then the logical volume Retrieving the last consecutive logical volume in “reverse order”, moving from the last LBA before capacity expansion of the last logical volume to the LBA after capacity expansion in reverse order, and repeating the logical volume movement process for all logical volumes It is characterized by moving from LBA before capacity expansion to LBA after capacity expansion.

  According to a second aspect of the invention, a computer that controls a storage apparatus in which a plurality of logical volumes are set on a RAID group composed of a plurality of physical disks is designated by specifying the capacity of each logical volume after the capacity expansion of the storage apparatus. A program for adding a disk and executing capacity expansion of a logical volume, the step of accepting the capacity specification of each logical volume after capacity expansion, and the logical volumes before and after capacity expansion in order from the logical volume set in the first LBA The step of obtaining the head and tail LBAs of the volume and registering them in the capacity expansion management table, and the logical volumes to which the head LBA and tail LBA do not move after capacity expansion are registered in the capacity expansion management table as “normal order”. Step and the first LBA does not move after capacity expansion, the last LBA A logical volume to be moved later is divided into logical volumes at positions where the LBA does not change before and after capacity expansion, and is registered in the capacity expansion management table with the first order as “normal” and the second order as “reverse”. The logical volume to which the first LBA and the last LBA move after the capacity expansion is registered in the capacity expansion management table as “reverse order”, and the logical volume in which the first LBA moves later and the last LBA does not move after capacity expansion. When a volume is added to the volume expansion management table by dividing the logical volume at a position where the LBA does not change before and after capacity expansion, registering the first half as "reverse order" and the second half as "normal order", and a physical disk Read the order of the first logical volume registered in the capacity expansion management table, and if the order is “Normal” Is the step of moving from the first LBA before capacity expansion of the logical volume to the LBA after capacity expansion, and if the order is “reverse order”, the logical volume from the next logical volume to the end of the “reverse order” logical volume And the process of moving the last logical volume from the last LBA before capacity expansion to the LBA after capacity expansion in reverse order, and repeating the logical volume movement processing to expand the capacity from the LBA before capacity expansion for all logical volumes And a step of moving to a later LBA.

  A third invention is a storage apparatus comprising a CPU, a memory, a physical disk group constituting a RAID group, and an area for accommodating a physical disk to be added. The memory has at least a capacity expansion management table in the CPU. A program for executing creation, a program for causing the CPU to execute capacity expansion processing, and a capacity expansion management table, wherein the capacity expansion management table creation program receives a capacity designation of each logical volume after capacity expansion; Obtaining the beginning and end LBA of each logical volume before and after capacity expansion and registering them in the capacity expansion management table in order from the logical volume set as the first LBA, and the beginning LBA and end LBA do not move after capacity expansion The logical volume is placed in the capacity expansion management table with the order of "normal order". For the logical volume in which the first LBA does not move after the capacity expansion and the last LBA moves after the capacity expansion, the logical volume is divided at a position where the LBA does not change before and after the capacity expansion. ”, A step of registering the latter half of the order in the capacity expansion management table as“ reverse order ”, and a step of registering the first LBA and the logical volume to which the end LBA moves after the capacity expansion as“ reverse order ”in the capacity expansion management table. For the logical volume in which the first LBA moves after the capacity expansion and the last LBA does not move later, the logical volume is divided at a position where the LBA does not change before and after the capacity expansion, and the first half is “reverse order” and the second half is “normal order”. To the capacity expansion management table, and the capacity expansion processing program A step of reading the order of the first logical volume registered in the capacity expansion management table; and a step of moving to the LBA after capacity expansion in order from the first LBA before capacity expansion of the logical volume if the order is “normal order”. If the order is "reverse order", the last of the consecutive logical volumes in the "reverse order" is searched from the next logical volume, and the LBA after capacity expansion in reverse order from the last LBA before capacity expansion of the last logical volume. And moving the logical volume to the LBA after capacity expansion from the LBA before capacity expansion for all logical volumes.

  According to the present invention, even when a storage apparatus in which a plurality of logical volumes are set on a RAID group composed of a plurality of physical disks is in operation, the capacity can be expanded by adding a physical disk to the storage apparatus. .

  FIG. 1 shows an example of a flow for creating a capacity expansion management table based on the capacity of each logical volume after capacity expansion designated prior to capacity expansion.

  In step S1 of FIG. 1, the storage apparatus accepts capacity designation of each logical volume after LDE.

  In step S2, it is confirmed whether or not there is a volume that has not been created in the capacity expansion management table in the logical volume on the RAID group. If there is no volume that has not been created, the process is terminated. .

  In step S3, the beginning and end LBAs of the logical volume before and after capacity expansion are calculated. In step S4, it is confirmed whether the head LBA moves later after capacity expansion. If it moves later, the process proceeds to step S8. When not moving, it progresses to step S5.

  In step S5, it is confirmed whether the data end LBA moves after capacity expansion. If not, the order is set to “normal order”. The process proceeds to step S11, and if it moves later, the process proceeds to step S7. Is obtained, the logical volume is divided at the position, the first half order is “normal order”, and the second half order is “reverse order”, and the process proceeds to step S11.

  In step S8, it is confirmed whether or not the data end LBA moves later after capacity expansion. If it moves later, the order is "reverse order", and the process proceeds to step S11. If it does not move later, the process proceeds to step S10. A position that does not change is obtained, the logical volume is divided at the position, the first order is set to “reverse order”, and the second order is set to “normal order”, and the process proceeds to step S11.

  In step S11, the beginning LBA, data end LBA, and order of the logical volume are registered in the capacity expansion management table, and the process returns to step S2.

  FIG. 2 shows an example of a flow for performing capacity expansion processing according to the generated capacity expansion management table. If the addition of the physical disk is confirmed in step S21, the process proceeds to step S22, where it is confirmed whether there is an unprocessed logical volume in the capacity expansion management table. If there is no unprocessed logical volume, the process ends. If there is, the process proceeds to step S23.

  In step S23, the order of the first logical volume among the unprocessed logical volumes on the capacity expansion management table is confirmed. If the order is not “normal order”, the process proceeds to step S25, and if the order is “normal order”. Advances to step S24, the logical volumes are rearranged in order from the LBA before capacity expansion to the LBA after capacity expansion from the top, and the process returns to step S22.

  In step S25, the order of the next logical volume is confirmed. If it is “reverse order”, the order of the next logical volume is further confirmed. If it is not “reverse order”, the process proceeds to step S26. Are rearranged from the LBA before capacity expansion to the LBA after capacity expansion up to the first logical volume in reverse order from the end, and the process returns to step S22.

  An example of a storage apparatus according to the present invention is shown in FIG. The storage apparatus according to the present invention includes at least a CPU 1 that controls the storage apparatus, a memory 2, a physical disk group 6 to 9 constituting a RAID group, and an area 10 that accommodates additional physical disks. At least a program 3 for causing the CPU 1 to create a capacity expansion management table, a program 4 for causing the CPU 1 to execute a capacity expansion process, and a capacity expansion management table 5 are held.

  FIG. 4 shows the capacity of each logical volume on a RAID group composed of physical disks P # 0, P # 1, P # 2, and P # 3 whose capacity is 1800 (hexadecimal 708) in LBA units. However, the physical disk P # 4 having the same capacity as the physical disk # 0 is added to the storage device in which the logical volume L # 0, L # 1, L # 2 of 2400 (hexadecimal 960) is set in LBA units. In the first embodiment, the capacities of the logical volume L # 1 and the logical volume L # 2 are expanded.

  When the storage apparatus accepts the capacity designation of each ethical volume after capacity expansion in the capacity expansion management table shown in FIG. 5, creation of the capacity expansion management table shown in FIG. 1 is started.

  In the case of the first embodiment, since the capacity of the logical volume L # 0 does not change, the data end LBA is changed from 599 (hexadecimal 257) to 479 (hexadecimal 1DF) by adding P # 4, and the processing order is “normal order”. "

  In the logical volume L # 1, the head LBA is 480 (hexadecimal 1E0), the data end LBA is 959 (hexadecimal 3BF), and the processing order is “normal”.

  However, the capacity of the logical volume L # 1 is expanded, and the last LBA becomes 1259 (hexadecimal 4EB).

  Therefore, in the logical volume L # 2, the first LBA after capacity expansion is 1260 (hexadecimal 4EC), and the data end LBA is 1739 (hexadecimal 6CB). That is, in the logical volume L # 2, the leading LBA moves backward and forward at the capacity expansion and the LBA at the end of the data moves forward, so that the LBA does not move before and after the capacity expansion. In this case, the LBA is 1500 (hexadecimal). 5DB), the first half is “L # 2-1”, the “reverse order” is the second half “L # 2-2”, and “normal order” is registered in the capacity expansion management table.

  When the capacity expansion management table is completed, the storage apparatus starts the capacity expansion process flow shown in FIG. When a physical disk is added, data rearrangement is performed according to the processing order of the capacity expansion management table shown in FIG.

  FIG. 6 shows the active capacity expansion result of the first embodiment, and shows the arrangement of each physical disk, each logical volume, and the order of processing before and after capacity expansion.

  FIG. 7 shows an example in which the present invention is applied to the storage device before capacity expansion used in the first embodiment and the example of causing data corruption (capacity expansion example of the first logical volume) shown in FIGS. 18 and 20 is processed. A capacity expansion management table created in the second embodiment is shown in FIG.

  In this case, the data end LBA of the logical volume L # 0 is the same as the capacity expansion management table shown in FIG. 5 and is changed from 599 (hexadecimal 257) to 480 (hexadecimal 1DF), and the processing order is “normal”. Become.

  In the logical volume L # 1, the head LBA is 840 (hexadecimal 348), the data end LBA is 1319 (hexadecimal 527), and the processing order is “reverse order”.

  In logical volume L # 2, the top LBA is 1320 (hexadecimal 528), the LBA at the end of data is 1800 (hexadecimal 707), and the processing order is “reverse order”.

  In this example, the logical volume is not divided.

  When the capacity expansion management table is completed, the storage apparatus starts the capacity expansion process flow shown in FIG. When a physical disk is added, data rearrangement is performed according to the processing order of the capacity expansion management table shown in FIG.

  FIG. 9 shows the capacity expansion result of the second embodiment, and shows the arrangement of each physical disk, each logical volume, and the processing order before and after capacity expansion.

Capacity expansion management table creation flow example Example of capacity expansion processing flow Examples of storage devices according to the present invention Example 1 Capacity expansion management table of the first embodiment Capacity expansion result of Example 1 Example 2 Capacity expansion management table of the second embodiment Capacity expansion result of Example 2 Conventional example 1 (before capacity expansion) Conventional example 1 (after capacity expansion) Conventional example 2 (before capacity expansion) Conventional example 2 (after capacity expansion) Restriction condition of Conventional Example 2 Capacity expansion example (equal expansion) Capacity expansion example (addition of logical volume) Capacity expansion example (last logical volume capacity expansion) Capacity expansion example (capacity expansion of the first logical volume) Capacity expansion example (multiple logical volume capacity expansion) Data destruction example due to capacity expansion

Explanation of symbols

1 CPU
2 Memory 3 Capacity Expansion Management Table Creation Program 4 Capacity Expansion Processing Program 5 Capacity Expansion Management Table 6, 7, 8, 9 Disk Group Constructing RAID Group 10 Area Accommodating Additional Physical Disk

Claims (3)

A method for expanding the capacity of a logical volume by specifying the capacity of each logical volume after capacity expansion and adding a physical disk to a storage device in which a plurality of logical volumes are set on a RAID group composed of a plurality of physical disks Because
Accept capacity specification of each logical volume after capacity expansion,
Obtain the beginning and end LBA of each logical volume before and after capacity expansion in order from the logical volume set in the first LBA, and register it in the capacity expansion management table.
For the logical volume whose first LBA and last LBA do not move after capacity expansion, register the order as “normal” in the capacity expansion management table.
For logical volumes to which the first LBA does not move after capacity expansion and the last LBA move later, the logical volume is divided at a position where the LBA does not change before and after capacity expansion, and the first half order is “normal order” and the second half order. Is registered in the capacity expansion management table as "reverse order"
The logical volume to which the first LBA and the last LBA move later after capacity expansion is registered in the capacity expansion management table as “reverse order”.
For logical volumes where the first LBA moves after capacity expansion and the last LBA does not move later, the logical volume is divided at a position where the LBA does not change before and after capacity expansion, with the first half as “reverse order” and the second half as “normal order”. Register in the capacity expansion management table,
When a physical disk is added, read the order of the first logical volume registered in the capacity expansion management table,
When the order is “normal order”, the logical volume is moved from the first LBA before capacity expansion to the LBA after capacity expansion in order,
When the order is “reverse order”, the end of the consecutive logical volumes in the “reverse order” is searched from the next logical volume and moved to the LBA after capacity expansion in reverse order from the last LBA before capacity expansion of the last logical volume. And
A logical volume capacity expansion method, wherein the logical volume movement process is repeated to move all logical volumes from an LBA before capacity expansion to an LBA after capacity expansion. Add a physical disk by specifying the capacity of each logical volume after expanding the capacity of the storage device to a computer that controls a storage device in which multiple logical volumes are set on a RAID group consisting of multiple physical disks. A program for expanding the capacity of a volume,
Receiving a capacity specification of each logical volume after capacity expansion;
Obtaining the head and tail LBA of each logical volume before and after capacity expansion in order from the logical volume set in the head LBA and registering it in the capacity expansion management table;
A logical volume in which the first LBA and the last LBA are not moved after the capacity expansion is registered in the capacity expansion management table in the order of “normal order”;
For logical volumes to which the first LBA does not move after capacity expansion and the last LBA move later, the logical volume is divided at a position where the LBA does not change before and after capacity expansion, and the first half order is “normal order” and the second half order. Registering in the capacity expansion management table as "reverse order",
A logical volume to which the first LBA and the last LBA move later after capacity expansion is registered in the capacity expansion management table as “reverse order”;
For logical volumes where the first LBA moves after capacity expansion and the last LBA does not move later, the logical volume is divided at a position where the LBA does not change before and after capacity expansion, with the first half as “reverse order” and the second half as “normal order”. A step of registering in the capacity expansion management table;
A step of reading the order of the first logical volume registered in the capacity expansion management table when a physical disk is added;
When the order is “normal order”, the logical volume is moved from the first LBA before capacity expansion to the LBA after capacity expansion in order,
When the order is “reverse order”, the end of the consecutive logical volumes in the “reverse order” is searched from the next logical volume and moved to the LBA after capacity expansion in reverse order from the last LBA before capacity expansion of the last logical volume. Step to
A logical volume capacity expansion program comprising: repeating the logical volume movement process to move all logical volumes from an LBA before capacity expansion to an LBA after capacity expansion. A storage device comprising a CPU, memory, a group of physical disks constituting a RAID group, and an area for accommodating additional physical disks,
The memory holds at least a program for causing the CPU to create a capacity expansion management table, a program for causing the CPU to perform capacity expansion processing, and a capacity expansion management table.
Capacity expansion management table creation program
Receiving a capacity specification of each logical volume after capacity expansion;
Obtaining the head and tail LBA of each logical volume before and after capacity expansion in order from the logical volume set in the head LBA and registering it in the capacity expansion management table;
A logical volume in which the first LBA and the last LBA are not moved after the capacity expansion is registered in the capacity expansion management table in the order of “normal order”;
For logical volumes to which the first LBA does not move after capacity expansion and the last LBA move later, the logical volume is divided at a position where the LBA does not change before and after capacity expansion, and the first half order is “normal order” and the second half order. Registering in the capacity expansion management table as "reverse order",
A logical volume to which the first LBA and the last LBA move later after capacity expansion is registered in the capacity expansion management table as “reverse order”;
For logical volumes where the first LBA moves after capacity expansion and the last LBA does not move later, the logical volume is divided at a position where the LBA does not change before and after capacity expansion, with the first half as “reverse order” and the second half as “normal order”. Registering in the capacity expansion management table,
Capacity expansion processing program
A step of reading the order of the first logical volume registered in the capacity expansion management table when a physical disk is added;
When the order is “normal order”, the logical volume is moved from the first LBA before capacity expansion to the LBA after capacity expansion in order,
When the order is “reverse order”, the end of the consecutive logical volumes in the “reverse order” is searched from the next logical volume and moved to the LBA after capacity expansion in reverse order from the last LBA before capacity expansion of the last logical volume. Step to
And a step of repeating the logical volume migration processing to migrate all logical volumes from an LBA before capacity expansion to an LBA after capacity expansion.

Images