JP3785127B2 - Disk array control device and data writing method in disk array control device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a disk array control device of a computer system and a data writing method in such a disk array control device.
[0002]
[Prior art]
Conventionally, as one of the speed-up methods of RAID (Redundant Array of Independent Disks), log writing (log structured file system) is performed in cluster units for all user areas of a logical disk to be controlled.
[0003]
In this method, in particular, in response to a plurality of random write requests for RAID5, they are continuously written to the physical disk in a cluster size. As a result, it is possible to eliminate the need to read data before write processing (original data and parity value before writing), which was necessary for conventional recalculation of parity, and realize efficient write processing. Was. Such techniques are disclosed in, for example, Japanese Patent Application Laid-Open Nos. 11-53235, 2001-184172, and 2002-14776.
[0004]
[Problems to be solved by the invention]
However, in such a RAID acceleration system, it is necessary to perform address conversion using the address conversion table for the data block management units of all user areas of the target logical disk. For this reason, an address conversion table whose size is proportional to the capacity of the logical disk to be controlled is required. When the size of the disks constituting the logical disk is large or when the number of disks is large, the capacity of the logical disk increases. The address translation table could exceed the memory capacity that the RAID controller can use.
[0005]
In order to avoid this problem, for example, the following two methods are conceivable.
[0006]
1. Place the address translation table on the disk, load and access only a subset of the address translation table that contains the currently referenced entry, and write the modified subset of the address translation table back to the disk at the appropriate time ( Address translation table caching). In this way, a large address conversion table is managed with a limited memory capacity.
[0007]
2. Limit the capacity of the logical disk to a size that can be controlled by the size of the address translation table that fits in the available memory.
[0008]
However, the method (1) has the advantage that the size of the address conversion table is not limited and the size of the logical disk to be controlled need not be restricted. Restoration takes a certain amount of time, and if the processing is not completed, address determination cannot be performed, so I / O processing for the original user data also stops, eventually increasing turnaround time and causing performance degradation It can be.
[0009]
The method (2) does not have the above-mentioned disadvantages in terms of performance, but instead has the disadvantage that the capacity of the logical disk is limited by the size of the memory for storing the address conversion table.
[0010]
The present invention provides a disk array control device capable of realizing high-speed data writing without imposing restrictions on the capacity of a logical disk to be controlled while keeping the amount of memory used for an address conversion table constant. Objective.
[0011]
[Means for Solving the Problems]
  In order to achieve the above object, the first invention of the present invention provides:In a disk array control device for controlling a disk array provided with at least one logical disk, in one logical diskA physical cluster management table that manages the physical cluster status of the disk array;In the one logical diskAn address conversion table for storing a logical block number and a physical block number corresponding to the logical block number in association with each other, and whether a physical block number of a data block to be written is registered in the address conversion table. When it is determined that the determination unit and the physical block number of the data block to be written are registered in the address conversion table, a free physical cluster may be secured by referring to the physical cluster management table. When it is determined that a means for determining whether or not a free physical cluster can be secured, the free physical cluster is secured, and the data block to be written is written to the secured free physical cluster. Means that can be accessed without the need for address translation. Direct address cluster consisting capacitor blocks only, a mix of the indirect address cluster containing data blocks that require address translation1 logical diskA disk array control device, characterized in that
[0012]
  According to the second aspect of the present invention, the means for determining whether or not there is a free physical cluster to be written next, or whether or not there is a free physical block in the physical cluster to be written next. And a means for determining whether or not an indirect address data block written at a physical address on a disk different from the requested logical address exists in the write candidate logical block when it is determined that the logical block is present. And a means for selecting an indirect address data block belonging to a physical cluster having the smallest number of effective blocks when it is determined that the indirect address data block exists in the logical block as the write candidate. Write the indirect address data block that has been written as the data block to be written The disk array controller according to claim 1, wherein a.
  Further, according to the third aspect of the present invention, a physical cluster management table for managing the state of a physical cluster in one logical disk, a logical block number in the one logical disk, and a physical corresponding to the logical block number In a data writing method in a disk array control apparatus for controlling a disk array provided with at least one logical disk having an address conversion table for storing a block number in association with the physical number of a data block to be written It is determined whether or not a block number is registered in the address conversion table, and when it is determined that a physical block number of the data block to be written is registered in the address conversion table, the physical cluster management table Refer to free physical It is determined whether or not a raster can be secured, and if it is determined that a free physical cluster can be secured, the free physical cluster is secured, and the write is made to the secured free physical cluster. A logical disk comprising a step of writing a data block, in which a direct address cluster consisting only of data blocks that can be accessed without requiring address translation and an indirect address cluster including data blocks that require address translation are mixed. Is a data writing method in a disk array control device.
[0013]
DETAILED DESCRIPTION OF THE INVENTION
First, definitions of terms used in the embodiment of the present invention will be described.
[0014]
1. Stripe set
In RAID, when configuring a logical disk from a plurality of disks, the address space of the logical disk is allocated to the physical disks in order for each fixed size by a method called striping. This fixed size data is called a stripe.
[0015]
In particular, in the case of RAID5, data stripes and a corresponding parity stripe are combined into one set to generate parity data and perform data reproduction processing in the event of any one disk failure.
[0016]
On the RAID5 logical disk, the continuous area is the size of "the number of disks of the stripe size managed by the RAID controller (data is the number of disks of the stripe size-1"), and each of the included areas A group of stripes corresponding to a parity stripe included in only one stripe size data of a disk is called a “stripe set”.
[0017]
For example, FIG. 13 shows a RAID 5 logical disk composed of three disks. S0, S1,... Indicate stripes and have a fixed length of 64 KB, for example. P0, P1,... Are parity blocks and have a stripe size. In this figure, “S0, S1, P0” is a stripe set. Similarly, “S2, P1, S3”, “P2, S4, S5” are stripe sets, respectively. On the logical disk, the arrangement of data stripes is S0, S1, S2, S3, S4, S5,..., And the parity stripe cannot be seen as data at the upper level.
[0018]
By writing a size that is an integral multiple of this size to RAID5 all at once, data read from the disk for parity recalculation can be omitted, and a significant performance improvement can be realized. This is one method for increasing the speed of RAID.
[0019]
2. cluster
Each data unit that divides the physical address space of the disk by stripe set size or an integral multiple of the stripe set, and by stripe set boundaries. In this method, data consisting of indirect address data blocks is written to the disk in units of “cluster”. Here, the direct address means that it can be accessed without requiring address conversion, and the indirect address means that it cannot be accessed without performing address conversion.
[0020]
3. Address translation table
In the RAID acceleration method, the block number when the file system of the host computer requests access is a logical block number and a virtual block number. This logical block number is associated with a physical block number (block number on the logical disk) by an “address conversion table” managed by the RAID controller.
[0021]
The address conversion table exists on the target logical disk, and an area different from the data part is allocated. The address conversion table is a table in which physical block numbers for logical block numbers are registered. When a new data block is written, the physical block number (= block number on the logical disk) to which the block is written is registered in the corresponding logical block number entry. Conversely, when referring to a data block, a value registered with respect to the logical address is obtained, and an actual address is obtained and referred to using the value as a block number on the logical disk.
[0022]
4). Data block management unit (block)
The RAID controller of the disk array control apparatus according to the embodiment of the present invention is a unit for managing data and has a fixed length (for example, 4 KB). If necessary, addresses are registered in the address conversion table and managed in this unit.
[0023]
5). Logical block number
The “logical block number” is a number obtained by converting an access request address of an I / O received by the RAID controller from a host computer or the like from a host computer or the like in a data block management unit managed by the RAID controller. In the RAID acceleration method, the block number when requested from the host computer is a logical block number, which is a virtual block number. This logical block number is associated with a physical block number (block number on the logical disk) by an “address conversion table” managed by the RAID controller. The byte offset value (address) on the logical disk is obtained from (physical block number) x (block size [Byte]).
[0024]
6). "Direct address data block"
In the embodiment of the present invention, a data block management unit that is not registered in the address translation table. The logical block number is used as the physical block number.
[0025]
That is, it is a block that is placed (written) on the disk according to the originally requested address.
[0026]
7). "Indirect address data block"
In the embodiment of the present invention, it is a data block management unit registered in the address conversion table. The address at the time of access is obtained from the physical block number (registered) corresponding to the logical address number on the address conversion table.
[0027]
That is, a block that is placed (written) at an address on the disk that is different from the originally requested address.
[0028]
8). Direct address cluster
It is a cluster composed of “direct address data blocks”.
[0029]
9. Indirect address cluster
A cluster composed of “indirect address data blocks”.
[0030]
10. Logical cluster
An address space used for data access from the upper level is called a “logical address space”. A data unit when a logical address space is divided into “cluster” units from the top is called a “logical cluster”.
[0031]
11. Physical cluster
A cluster on a logical disk with respect to a logical cluster is called a “physical cluster”. In the following description, “cluster” refers to “physical cluster”.
[0032]
12 Repack processing
A process for creating a free physical cluster by collecting valid blocks of a logging cluster.
[0033]
The disk array control apparatus according to the embodiment of the present invention is based on the following conditions.
[0034]
1. The RAID controller control algorithm is implemented as RAID controller firmware.
[0035]
2. A logical disk is configured by RAID.
[0036]
3. Write data for the RAID controller is converted by the high-speed module by registering its address in the address conversion table, and even if it is a series of write data blocks whose addresses are not continuous, it becomes a continuous address on the disk. Write the address as follows. The speed-up module is obtained by mounting the embodiment of the present invention in the form of a module as part of the firmware of the RAID controller. This speed-up module improves the processing performance of the RAID controller.
[0037]
4). The size written as a continuous address on the disk is a “stripe set (or an integer multiple thereof)” that can be efficiently written to the disk because it is not necessary to read the parity or original data on the disk.
[0038]
Hereinafter, a disk array control apparatus of a computer system according to an embodiment of the present invention will be described with reference to the drawings.
[0039]
FIG. 1 is a diagram showing a configuration of a computer system according to an embodiment of the present invention.
[0040]
As shown in the figure, a CPU 2 and a main memory 3 are connected to the CPU bus 1.
[0041]
The CPU 2 controls the entire computer system, and uses the main memory 3 as a work area. The CPU bus 1 is connected to the PCI bus 5 via the PCI-PCI bridge 4.
[0042]
A RAID controller PCI card 11 is connected to the PCI bus 5. The RAID controller PCI card 11 includes a PCI-PCI bridge 12, an internal bus 13, a cache memory 14, a SCSI controller 15, a dedicated CPU 16, a local RAM 17, and a flash ROM 18.
[0043]
The PCI-PCI bridge 12 performs bridge control between the PCI bus 5 and the internal PCI bus 13. A cache memory 14, a SCSI controller 15, and a dedicated CPU 16 are connected to the internal PCI bus 13. A local RAM 17 and a flash ROM 18 are connected to the dedicated CPU 16.
[0044]
The cache memory 14 temporarily stores data that the computer accesses to the disk, and is used for the purpose of improving the access speed. Alternatively, the RAID5 logical disk is also used as an area for parity calculation. For the control of the cache memory 14, for example, a management method such as set associative is generally used.
[0045]
The SCSI controller 15 is an interface for accessing a disk from the RAID controller PCI card 11. The dedicated CPU 16 instructs the SCSI controller 15 to transfer data between the cache memory 14 and each of the disks 21-1 to 21-n and transmit a SCSI command to each disk.
[0046]
The local RAM 17 becomes a work data area necessary for processing of the control program of the dedicated CPU 16. The flash ROM 18 stores a control program for the dedicated CPU 16. The SCSI controller 15 is connected to the disks 21-1 to 21-n via the SCSI bus 20.
[0047]
FIG. 2 is a diagram showing a configuration of the cache memory 14 of the computer system according to the embodiment of the present invention. This cache memory 14 is a cache memory of a RAID controller managed by the set associative method.
[0048]
As shown in the figure, the cache memory 14 has m sets and n ways. A cell indicates one cache block. Here, it is assumed that the size is 4 KB or 16 KB.
[0049]
The cache management table is a two-dimensional array of (m, n) like the cache memory. The entry contains (valid bit 31, update bit 32, other bit 33, tag 34). The valid bit 31 indicates whether the cache memory is valid / invalid (whether or not valid data is contained) by 1 bit, and the update bit 32 indicates whether or not the data in the cache memory is updated. In addition, there is a bit 33 used for other management (description omitted), and a data block number is set in the tag 34.
[0050]
When reading / writing data into the cache memory, the set number is obtained from the address according to the following equation.
[0051]
(Set number) = (Block number) mod (Number of sets m)
After determining the set number, the cache management table is referenced to check whether there is an empty entry in the set. If there is an empty entry, it is used for the entry in the cache management table corresponding to the entry. If there is no free entry, a cache block belonging to the same “column” is freed and secured there, or processing is performed without using a cache.
[0052]
FIG. 3 is a diagram showing the layout of the logical disk address space of the disk in the RAID acceleration method, (a) is a diagram showing the stripe arrangement on the physical disk, and (b) is the stripe arrangement on the logical disk. It is a figure which shows a layout.
[0053]
3, d0, d1,... Indicate stripes including data. The stripe has a size of about 64 KB, for example. Further, p0, p1, p2,... Are parity stripes of (d0, d1), (d2, d3), and (d4, d5), respectively.
[0054]
Here, (d0, d1, p0), (d2, p1, d3), and (p2, d4, d5) are “stripe sets”, respectively.
[0055]
In this example, it is assumed that two stripe sets are one cluster. Therefore, the physical cluster 0 is d0, d1, (p0,) d2, (p1,) d3, and the physical cluster 1 is (p2,) d4, d5, d6, d7 (, p3).
[0056]
The entire logical disk is divided into three areas: a “direct address area”, a “margin area”, and a “management data area”.
[0057]
The “direct address area” is secured by the size that can be used as a disk to be shown to the host (host computer). For example, when a capacity of 100 GB can be used as a logical disk, the size of the “direct address area” is 100 GB.
[0058]
The “margin area” is composed of a plurality of physical clusters. By repeating log writing of this method, a physical cluster in which valid data blocks are 100% clogged immediately after writing is written to the physical block number of the logical data block by successive indirect address physical cluster writing. As a result of moving to the position of, the proportion of valid blocks in the indirect address physical cluster is steadily decreasing. As a result, all the logical blocks cannot be accommodated only by the physical cluster in the direct address area. By preparing a physical cluster roughly as a margin area in advance, an empty physical cluster exists even when a log cluster is written, and an indirect address physical cluster can be written.
[0059]
However, the number of indirect address physical clusters will eventually increase, and there will be no empty clusters. In order to avoid this state, a physical block including valid data is collected when the load on the RAID controller is low, and a process of generating a free physical cluster is performed (repacking process).
[0060]
The “management data area” is an area for placing a data structure for managing the entire logical disk. This area includes an address conversion table and other management information. The address conversion table is loaded into a part of the cache memory of the RAID controller when the system is started, and written back to the disk when the system is stopped. Other management information is similarly loaded / saved as necessary.
[0061]
FIG. 4 is a diagram showing a layout of a “logical address space” corresponding to FIG.
[0062]
The logical address space is “a logical disk address space visible from outside the RAID controller”. For example, the host computer accesses the data by building a file system for the “logical address space”. Treat as a single disc.
[0063]
In the figure, when all the data blocks constituting the “logical cluster” are “direct address data blocks”, the “logical cluster number” matches the number of the “physical cluster” in which the data blocks are placed. Such clusters are called “direct address physical clusters”, “direct address logical clusters”, or simply “direct address clusters”.
[0064]
FIG. 5 is a diagram showing an address conversion table.
[0065]
When the logical cluster is an “indirect address cluster”, the data block management unit constituting the logical cluster is assigned to an address on the disk corresponding to a “physical block number” different from the “logical block number”.
[0066]
The “address conversion table” records and manages the correspondence between the “logical block number” and the “physical block number”. In the conventional RAID acceleration method, entries for all logical blocks are prepared in the address conversion table. On the other hand, in the embodiment of the present invention, the amount of memory that can be used in the address conversion table is limited, and it is assumed that not all logical blocks are registered in the address conversion table, but only a part can be registered. For a logical block that cannot be registered, the logical block number is treated as a physical block number.
[0067]
Therefore, only “data block management units” belonging to “indirect address clusters” are registered in the address conversion table.
[0068]
The address conversion table is registered in units of “indirect address logical clusters”. Conversely, data block management units belonging to the “direct address logical cluster” are not registered in the address conversion table.
[0069]
In the figure, it is assumed that one logical cluster is composed of 8 data blocks, and it becomes a table entry of three sets of “logical cluster number”, “logical block number”, and “physical block number”. ing. However, the “logical block number” item is not necessary. Instead, it can be obtained as follows (in order to save the amount of memory used by the address translation table, it is better to omit the logical block number).
[0070]
The main purpose of use of this table is to determine whether or not a specific “logical block number” is registered in the address translation table, and if registered, the “physical block number” assigned to that logical block number. Knowing the value of "number".
[0071]
Now, when there is a “logical block number” LBn, the logical cluster number is
"Logical cluster number"
A = (LBn) / "number of data blocks per logical cluster"
(However, “/” indicates integer division.)
The "block offset number in the logical cluster" is
"Logical cluster block offset number"
B = (LBn) mod "Number of data blocks per logical cluster"
(However, “mod” indicates modulo operation)
It becomes. Therefore, it is only necessary to obtain the physical block number registered in the cluster with the logical cluster number A by referring to the address conversion table.
[0072]
Next, the basic processing flow of the write operation of the disk array control apparatus according to the embodiment of the present invention will be described with reference to the flowchart of FIG.
[0073]
FIG. 6 shows a processing flow when one data block management unit LBi is written to the disk.
[0074]
For the explanation of FIG. 6, the situation as shown in FIGS. 7A to 7C is specifically considered. Hereinafter, for the sake of explanation, the number of blocks, the number of entries in the table, and the like are indicated by numbers far smaller than those of the actual system.
[0075]
FIG. 7A shows a physical cluster on the logical disk. Here, LB0, LB1,... Each indicate a logical block, and the logical block number is represented by a subscript.
[0076]
In this example, the physical cluster is composed of four data blocks. For example, the physical cluster 0 is composed of four data blocks LB0 to LB3.
[0077]
In this example, there are 16 physical clusters in the data area on the logical disk. However, the “logical address space” of the logical disk is as wide as 13 physical clusters, and this number of “logical clusters” exists.
[0078]
Note that direct address data blocks may exist in the range of physical clusters 0-12. This area is called a “direct address area”. The physical clusters 13 to 15 are “margin areas”. Only “indirect address data block” can be arranged in this portion.
[0079]
In FIG. 7A, all logical blocks are arranged at physical block number positions equal to the numbers. That is, all logical clusters are “direct address clusters”. Also, no valid data block is arranged in the margin area.
[0080]
FIG. 7B is a table for managing the states of all the physical clusters shown in FIG. 7A, and is placed in the cache memory of the RAID controller. The index is a physical cluster number. Possible values are
Direct address cluster ・ ・ ・ All blocks in the cluster are direct address data blocks
Indirect address cluster: One or more indirect address data blocks are included in the cluster.
Empty cluster: No valid block is included in the cluster
It is three.
[0081]
Here, the physical clusters 0 to 12 are “direct address clusters”, and the others are “empty clusters”.
[0082]
FIG. 7C shows an address conversion table, which is placed in the cache memory of the RAID controller. The number of the logical cluster of the indirect address and four physical block numbers where the logical blocks constituting the indirect address are registered are registered. In this state, nothing is registered in the address translation table (because there is no indirect address logical cluster). In this example, it is assumed that an address conversion table large enough to register blocks for four logical clusters is prepared.
[0083]
Here, the state when writing logical blocks to the disk in the order of LB9, LB6, LB8, LB0, LB17, LB15, LB2, LB7, LB6.
[0084]
In FIG. 6, the initial LBi is “LB9”. In step 601, since LB9 is not registered in the AMT (address translation table), the process proceeds to step 602.
[0085]
In step 602, an entry necessary for registering the “direct address logical cluster” 2 to which LB9 belongs exists in the address translation table. Therefore, the process proceeds to step 603, where “logical cluster” 2 (all blocks) is address translated. Register in the table. That is, “2” is registered in the first “logical cluster number” of the address conversion table and “8, 9, 10, 11” is registered in the physical block number (step 603).
[0086]
Proceeding to step 606, since there is no physical cluster PC currently being written, the process proceeds to step 607 to search for a free physical cluster. Here, the physical cluster PC is a cluster secured in step 609 and does not exist at the initial writing stage.
[0087]
From the physical cluster management table shown in FIG. 7B, the physical cluster 13 is found to be an “empty cluster” and is selected as a physical cluster PC. Address cluster "(steps 608 and 609).
[0088]
In step 612, the first block (physical block number 52) of the physical cluster 13 is assigned to LB9, and in step 613, "52" is registered in the entry of logical block number 9 in the address translation table (AMT). In FIG. 6, PBnext means a pointer for writing.
[0089]
In step 614, since three empty blocks still remain in the physical cluster 13, the process proceeds to step 616, and actual writing is not yet performed. For example, as shown in FIG. 8D, a write buffer for one cluster is prepared in the memory, and data of logical blocks determined to be written are placed here, and all blocks to be written to the physical cluster in step 614 are stored. Is determined, the data is collectively written to the physical cluster in step 615.
[0090]
FIG. 8 shows a state where the processing has progressed so far. FIG. 8A is a diagram of a physical cluster, and there is no latest data of LB9 already at the position of physical block number 9 where LB9 was initially placed.
[0091]
The newly written LB9 data is reserved to be written in the first block of the physical cluster 13, and the actual data is stored in the write buffer as shown in FIG. 8D. In the physical cluster management table shown in FIG. 8B, the physical clusters 2 and 13 are described as “indirect address clusters”, and the logical cluster 2 is registered in the address conversion table shown in FIG. In particular, 52, which is the number of the first block of the physical cluster 13, is associated with the physical block number of the write target LB9.
[0092]
Similarly, FIG. 9 shows a state when logical blocks are processed in the above order (LB9, LB6, LB8, LB0, LB17, LB15, LB2, LB7, LB6...) According to the flowchart of FIG. In FIG. 9 (a), the portions marked with x indicate invalid (out of date) data by writing. LB15 indicates that the address conversion table is overwritten at the position of the current physical block number because there is no empty space in step 602.
[0093]
In the embodiment of the present invention, performing new writing as it is at the position of the physical block number where the current logical data block is placed is called “overwriting”.
[0094]
Further, the writing in the above order (LB9, LB6, LB8, LB0, LB17, LB15, LB2, LB7, LB6...) Is a writing pattern called random writing. In other words, the logical addresses are written in a different order, and normally writing to the disk in this order causes a seek, resulting in poor efficiency.
[0095]
The portions written on the disk at the time of FIG. 9A are physical clusters 13 and 14. Since writing is performed in units of stripe sets (multiples of each), reading for parity calculation does not occur and the seek range on the disk is narrow, so writing can be performed very efficiently and at high speed.
[0096]
In the present embodiment, the total number of logical data blocks arranged as an indirect address cluster is limited within the range of the total number of entries in the address translation table.
[0097]
In other words, batch writing as an indirect address cluster is possible within a range that can be registered in a fixed-size address conversion table, so that it is not necessary to limit the capacity of the target logical disk with the size of the address conversion table as in the prior art.
[0098]
If a free physical cluster cannot be secured in step 608, LBi is overwritten and written to the address of the physical block where the current logical data block is located (step 610). If it is determined in S602 that there are not enough entries in the address conversion table, the logical block LBi to be written is directly written as an address data block (step 604).
[0099]
In steps 612 and 613, data is not yet written at the address of the physical block number on the disk, and remains on the write buffer. Therefore, when reading these data, it is necessary to refer to the write buffer until it is written to the disk.
[0100]
Regarding the write buffer, in this example, the memory for one physical cluster is secured and the write data is placed. However, when the data to be written exists on the cache memory, it is managed as a pointer string to the cache memory. May be. Such an implementation can eliminate the overhead of data copy.
[0101]
FIG. 10 is a flowchart for explaining the overall operation of the writing process of the disk array control apparatus according to the embodiment of the present invention.
[0102]
This processing is, for example, logic applied to write processing to a disk of the dirty cache of the RAID controller. Specifically, consider a process of periodically writing dirty data from a RAID controller cache to a disk. This is a situation in which any dirty data on the cache has to be written, and is repeatedly executed until there is no need to write (for example, a sufficiently free cache block is generated). Note that the processing of FIG. 6 is used for writing one basic data block.
[0103]
First, let us consider a case where there is no free physical cluster and there is no free physical block in the physical cluster currently being written out.
[0104]
In this case, the following is considered.
[0105]
When there is no “free cluster” and “there is no physical cluster to be written out”, writing as a logging cluster cannot be performed. For this reason, even if any logical block is selected as a write target, the only choice is to write (overwrite) the current physical block assigned to that logical block.
[0106]
However, there is a possibility that an empty physical cluster will eventually occur. For example, there is a possibility that the repacking process is proceeding at the same time, or the load is reduced and the repacking process starts preferentially and an empty physical cluster may be generated.
[0107]
When free physical clusters occur, it is advantageous to write indirect address data blocks preferentially in the form of logged clusters. This is because even if the indirect address data block is written in the log cluster, the entry of the “address conversion table” is not consumed again (because it is already registered). Therefore, if a direct address data block exists as a write candidate, the direct address data block should be preferentially written by overwriting.
[0108]
Based on the above idea, in S801, it is determined whether there is a free physical cluster or there is a free physical block in the target physical cluster that is currently written, and if it is determined that it does not exist, In step S809, overwriting is attempted.
[0109]
FIG. 11 is a flowchart for explaining the operation of the overwrite writing process.
[0110]
As shown in the figure, first, in step S901, it is determined whether or not there is a direct address data block as a write candidate (S901). If it is determined in S901 that there is a direct address data block, the longest one of the direct address data block columns with consecutive addresses is selected for the direct address data block to be written (S902). Then, the column of the selected direct address data block is overwritten and written (S903). In S902, the column of the block with the longest direct address is selected because it is more efficient to write in a large size even if it is an overwrite. At this time, it is most efficient to write data at the stripe set boundary and stripe size.
[0111]
In this way, by delaying the writing of the indirect address data block, it becomes possible to write data in the logged stripe format at high speed without consuming entries in the address translation table when a free physical cluster occurs.
[0112]
On the other hand, let us consider a case where there is no direct address data block to be written in step 901.
[0113]
In this case, the following is considered.
[0114]
1. An indirect address data block can only be selected for writing.
[0115]
2. A certain amount of time is required for the writing process.
[0116]
3. During the writing process, the physical stripe to be written cannot be the target of the repack process.
[0117]
4). If an indirect address data block belonging to a physical cluster to be repacked is selected as a write target here, the efficiency of the repacking process deteriorates.
[0118]
5). Since there are currently few free physical clusters, we do not want the repacking process to be inefficient.
[0119]
6). A physical cluster to be repacked is a physical cluster with a small number of valid data blocks (in repacking, it is more efficient to merge physical stripes with a small number of valid blocks to create a free physical stripe).
[0120]
In consideration of the above circumstances, a block belonging to the physical stripe that contains the most effective blocks is selected from the indirect address data blocks.
[0121]
That is, if it is determined in S901 that no direct address data block exists, an indirect address data block belonging to a physical cluster having the largest number of valid blocks is selected from candidate indirect address data blocks (S904). The selected indirect address data block is overwritten (S905).
[0122]
By proceeding in this way, when a physical cluster with few valid blocks is selected for repacking, there are fewer opportunities for writing to that physical cluster, so processing can be continued and repacking efficiency is improved. .
[0123]
Consider a case where there is a free physical cluster in S801 in FIG. 10 or there is a free physical block in the physical cluster to be written out.
[0124]
In this case, the following is considered.
[0125]
1. There is a possibility to select the export method as a logging cluster.
[0126]
2. In particular, the indirect address data block does not need to secure a new entry in the address conversion table, so a writing method as a logging cluster is possible.
[0127]
3. In particular, a direct address data block can be written as a log cluster if a new address translation table entry can be secured.
[0128]
4). Address translation table entries are finite resources and should not be used as much as possible.
[0129]
In consideration of the above situation, the indirect address data block should be given priority as a log cluster write candidate. As a result, it is possible to proceed with processing so that consumption of entries in the address translation table does not increase.
[0130]
Further, by writing out the indirect address data block belonging to the physical cluster having the smallest number of effective blocks, the block moves to another physical cluster, and the number of effective blocks of the physical stripe having the small number of effective blocks can be further reduced. This has the meaning of improving the efficiency of repacking.
[0131]
Therefore, the indirect address data block belonging to the physical cluster having the smallest number of effective blocks should be a candidate for writing. As a result, the number of physical clusters with a small number of effective blocks increases, and the repacking efficiency can be improved.
[0132]
That is, in S801, if there is a free physical cluster, or if there is a free physical block in the current write target physical cluster, it is determined whether there is an indirect address data block in the write candidate block (S802). ).
[0133]
If it is determined in S802 that the write candidate block has an indirect address data block, an indirect address data block belonging to the physical cluster having the smallest number of effective blocks is selected (S810). Then, the indirect address data block is written to the log stripe by the write process of FIG. 6 (S811).
[0134]
Next, consider a case where there is no indirect address data block as a write candidate.
[0135]
In this case, the following is considered.
[0136]
1. The block to be written out is a direct address data block.
[0137]
2. When the address data block is directly written out in the logged cluster format, the time required is short, but on the other hand, the address translation table entry is consumed.
[0138]
3. Furthermore, since the direct address data block becomes an indirect address data block, even if a series of blocks having consecutive logical block numbers are discontinuously arranged on the disk as indirect addresses, the subsequent sequential access performance deteriorates. There is a disadvantage that.
[0139]
4). Therefore, when the direct address data block can be efficiently written by overwriting, it is more efficient to write by overwriting as much as possible.
[0140]
5). In particular, when a series of direct address data blocks having consecutive block numbers (= addresses) exists as a write target, the data can be written by overwriting relatively efficiently.
[0141]
Therefore, if there are more than a certain number of continuous direct address data blocks, the direct address data blocks should be written in bulk at the time of overwriting.
[0142]
That is, if it is determined in S802 that the indirect address data block does not exist in the write candidate block, the longest one of the direct address data block columns having consecutive addresses is selected for the direct address data block to be written. BL is set, and the number of blocks included in the column is set to N (S803).
[0143]
Next, in S804, it is determined whether or not the number of blocks N is greater than a predetermined value “A”. Here, “A” is set as a block length reference value, and this value designates a block length that can be overwritten at a relatively high speed by actual measurement. Alternatively, an implementation in which a value is determined from the state of the entire system as a dynamic parameter is possible.
[0144]
If it is determined in S804 that the number of blocks N is larger than the predetermined value “A”, the selected BL is collectively written to the current address (S813). In S813, writing may be performed at the stripe set boundary and stripe size, and in this case, the efficiency is further improved.
[0145]
By performing such processing, it is possible to prevent unnecessary consumption of entries in the address conversion table. Moreover, the effect of preventing the sequential performance from deteriorating by increasing the indirect address unnecessarily can be obtained.
[0146]
Next, the case where it is determined in S804 that there are no continuous direct address data blocks equal to or greater than the number A will be described. In this case, since efficient direct address data block overwriting cannot be expected, the following is considered.
[0147]
1. Evaluate the urgency of writing, and if it is necessary to write as soon as possible, write in a logging cluster that can be written at high throughput.
[0148]
2. As the degree of “emergency”, the dirty rate of the cache can be considered. The higher the Dirty rate, the sooner it needs to be exported.
[0149]
3. When the urgency is low, the address data block is overwritten directly in order to avoid an increase in indirect address data blocks (to save address translation table entries and prevent deterioration of sequential performance). In particular, the more indirect address data blocks already exist, the more direct address data blocks are overwritten.
[0150]
Based on the above considerations, in S805, one of the write candidate direct address data blocks is selected to determine whether to write to the logged cluster or to overwrite. FIG. 12A is a flowchart for explaining the processing in S805.
[0151]
As shown in the figure, one of the direct address data blocks to be written belonging to the cache column (cache set) with the highest usage rate is selected (S1001). Next, the writing method is determined using the determination function f (α, β).
[0152]
FIG. 12B is a diagram for explaining the determination function f (α, β). In the figure, α means the dirty rate of the entire cache, and β means the usage rate (%) of the entire address translation table.
[0153]
In this example, when the dirty rate α of the cache is 90% or more, it is determined that “write to log cluster”. If it is less than that, determination is made by comparing α with the usage rate β (ratio of the number of entries in use to the total number of entries). The higher the dirty rate, or the smaller the number of indirect address data blocks, the more it is determined that “write to log cluster”. By changing the logic of the determination function f () according to the situation and implementation, the determination according to the actual situation can be performed.
[0154]
If it is determined in S1002 that the data is not overwritten, an instruction to write to the logging cluster is determined (S1003, 1005). On the other hand, if it is determined in S1002 that the data is overwritten, overwriting is determined (S1003, S1004).
[0155]
As a result, by the above method, when the dirty usage rate of the cache is low, it is possible to increase the number of indirect address data blocks and control so that the sequential performance is not deteriorated.
[0156]
Thereafter, in S806, it is determined whether or not it is decided to write to the log stripe (S806). If it is judged that it is not decided, the selected direct address data block is written overwritten ( S808). On the other hand, if it is determined, the direct address data block selected in the write processing of FIG. 6 is written. At this time, if possible, write in the logging cluster.
[0157]
Therefore, according to the disk array control apparatus of the embodiment of the present invention, in the disk system adopting the log format data block management method, the log format is limited within the usable range while limiting the capacity of the address conversion table. By managing this data block, it is possible to avoid the capacity limitation of the target logical disk.
[0158]
Further, when the urgency is high, the data is written in the log format, and when the urgency is low, the data is written in the non-log format, thereby avoiding an increase in the log data and avoiding the deterioration of the sequential access performance of the disk array control device.
[0159]
Note that the present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the scope of the invention at the stage of implementation. In addition, the embodiments may be appropriately combined as much as possible, and in that case, the combined effect is obtained. Furthermore, the above embodiments include inventions at various stages, and various inventions can be extracted by appropriately combining a plurality of disclosed constituent elements. For example, when an invention is extracted by omitting some constituent elements from all the constituent elements shown in the embodiment, when the extracted invention is implemented, the omitted part is appropriately supplemented by a well-known common technique. It is what is said.
[0160]
【The invention's effect】
As described above in detail, according to the present invention, the amount of memory used for the address conversion table is kept constant, the logical disk capacity to be controlled is not restricted, and the high speed of the conventional RAID acceleration method is used. Writing can be realized.
[Brief description of the drawings]
FIG. 1 is a diagram showing a configuration of a computer system according to an embodiment of the present invention.
FIG. 2 is a diagram showing a configuration of a cache memory 14 of the computer system according to the embodiment of the present invention.
FIG. 3 is a diagram showing a layout of logical disk address space of a disk in a RAID acceleration method, (a) is a diagram showing a stripe arrangement on a physical disk, and (b) is a stripe arrangement on a logical disk. It is a figure which shows a layout.
4 is a diagram showing a layout of a “logical address space” corresponding to FIG. 3; FIG.
FIG. 5 is a diagram illustrating an address conversion table.
FIG. 6 is a flowchart for explaining a basic processing flow of a write operation of the disk array control apparatus according to the embodiment of the present invention;
7A is a diagram showing a state of physical clusters on a logical disk, FIG. 7B is a table for managing the states of all physical clusters shown in FIG. 7A, and FIG. ) Is a diagram showing an address conversion table.
8A is a diagram showing a state of physical clusters on a logical disk, FIG. 8B is a table for managing the states of all physical clusters shown in FIG. 8A, and FIG. ) Is a diagram showing an address conversion table, and (d) is a diagram showing a write buffer.
9A is a diagram showing a state of physical clusters on a logical disk, FIG. 9B is a table for managing the states of all physical clusters shown in FIG. 9A, and FIG. ) Is a diagram showing an address conversion table, and (d) is a diagram showing a write buffer.
FIG. 10 is a flowchart for explaining the overall operation of the writing process of the disk array control apparatus according to the embodiment of the present invention;
FIG. 11 is a flowchart for explaining the operation of overwrite writing processing;
12A is a flowchart for explaining the processing in S805, and FIG. 12B is a diagram for explaining a determination function f (α, β).
FIG. 13 is a diagram showing a RAID 5 logical disk composed of three disks.
[Explanation of symbols]
1 ... CPU bus,
2 ... CPU,
3 ... main memory,
4 ... PCI-PCI bridge,
5 ... PCI bus,
11 ... RAID controller PCI card,
12 ... PCI-PCI bridge,
13 ... Internal PCI bus,
14 ... cache memory,
15 ... SCSI controller,
16 ... Dedicated CPU,
17 ... Local RAM,
18 ... Flash ROM,
20 ... SCSI bus,
21-1 to 21-n... Disk devices.

Claims (20)

In a disk array control device for controlling a disk array provided with at least one logical disk,
A physical cluster management table for managing the state of the physical cluster in one logical disk ;
An address conversion table for storing a logical block number in the one logical disk and a physical block number corresponding to the logical block number in association with each other;
Means for determining whether a physical block number of a data block to be written is registered in the address conversion table;
If it is determined that the physical block number of the data block to be written is registered in the address conversion table, whether or not a free physical cluster can be secured by referring to the physical cluster management table is determined. Means to judge,
When it is determined that a free physical cluster can be secured, the free physical cluster is secured, and a means for writing the data block to be written into the secured free physical cluster is provided,
A disk characterized by managing one logical disk by mixing a direct address cluster consisting only of data blocks that can be accessed without requiring address translation and an indirect address cluster containing data blocks that require address translation. Array controller.   A means for writing the data block to be written to a physical block address corresponding to the data block to be written that is registered in the address conversion table when it is determined that a free physical cluster cannot be secured; The disk array control apparatus according to claim 1, further comprising: When it is determined that the physical block number of the data block to be written is not registered in the address conversion table, an entry for registering the logical cluster to which the data block to be written belongs belongs to the address conversion table. Means for determining whether or not
When it is determined that there is an entry for registering the logical cluster to which the data block to be written belongs in the address conversion table, all the physical block numbers of the logical cluster to which the data block to be written belongs are all the addresses Means for registering in the conversion table;
Means for registering a physical block number of a write-destination data block in the empty physical cluster as a physical block number registered for the data block to be written newly registered in the address conversion table; 2. The disk array control apparatus according to claim 1, further comprising: When the logical cluster to which the data block to be written belongs is registered in the address conversion table, the physical block number corresponding to the logical cluster number to which the data block registered in the address conversion table belongs is used. And access
The access is performed using a logical block number of the data block as a physical block number when a logical cluster to which the data block to be written belongs is not registered in the address conversion table. 2. The disk array control device according to 1.   2. The disk array control apparatus according to claim 1, wherein the total number of logical data blocks arranged in the indirect address cluster is limited within a range of the total number of entries in the address conversion table. Means for determining whether there is a free physical cluster to be written next, or whether there is a free physical block in the physical cluster currently being written;
Means for determining whether or not an indirect address data block written to a physical address on a disk different from the requested logical address exists in the write candidate logical block when it is determined to exist;
Means for selecting an indirect address data block belonging to a physical cluster having the smallest number of effective blocks when it is determined that the indirect address data block exists in the logical block of the write candidate;
2. The disk array control apparatus according to claim 1, wherein the selected indirect address data block is written as the data block to be written.   When it is determined that there is no free physical cluster, or there is no free physical block in the current write target physical cluster, the write target block is registered in the address conversion table as the write target block. 7. The disk array control apparatus according to claim 6, further comprising means for writing to a physical block address corresponding to the data block to be written. If it is determined that the indirect address data block written to the physical address on the disk different from the requested logical address does not exist in the write candidate logical block, the address among the direct address data blocks of the write candidate Means for selecting the longest sequence of direct address data blocks that are consecutive,
7. The disk array according to claim 6, further comprising means for writing a column of the selected direct address data block when the column of the selected direct address data block satisfies a predetermined condition. Control device.   When it is determined that the predetermined condition is not satisfied, the dirty rate of the cache memory that temporarily stores data to be written to the disk array and the use of the address conversion table for each direct address data block of the write candidate 9. The disk array control apparatus according to claim 8, further comprising means for determining a writing method based on the rate.   The direct address data block as the write candidate is selected in order from a direct address data block belonging to a cache column having a high usage rate when it is determined that the predetermined condition is not satisfied. 9. The disk array control device according to 9. A physical cluster management table for managing the state of the physical cluster in one logical disk, the logical block number in the one logical disk, the address conversion table for storing correspondence between the physical block number corresponding to the logical block number In a data writing method in a disk array control device for controlling a disk array provided with at least one logical disk comprising :
Determine whether the physical block number of the data block to be written is registered in the address conversion table,
If it is determined that the physical block number of the data block to be written is registered in the address conversion table, whether or not a free physical cluster can be secured by referring to the physical cluster management table is determined. Judgment
If it is determined that a free physical cluster can be secured, the method comprises the steps of securing the free physical cluster and writing the data block to be written to the secured free physical cluster,
A disk characterized by managing one logical disk by mixing a direct address cluster consisting only of data blocks that can be accessed without requiring address translation and an indirect address cluster containing data blocks that require address translation. Data writing method in array controller.   When it is determined that a free physical cluster cannot be secured, the step of writing the data block to be written to a physical block address corresponding to the data block to be written registered in the address conversion table is further included. 12. The data writing method in the disk array control apparatus according to claim 11, further comprising: When it is determined that the physical block number of the data block to be written is not registered in the address conversion table, an entry for registering the logical cluster to which the data block to be written belongs belongs to the address conversion table. To determine whether or not
When it is determined that there is an entry for registering the logical cluster to which the data block to be written belongs in the address conversion table, all the physical block numbers of the logical cluster to which the data block to be written belongs are all the addresses Register in the conversion table,
A step of registering a physical block number of a write-destination data block in the free physical cluster as a physical block number registered for the data block to be written newly registered in the address conversion table; 12. The data writing method in the disk array control apparatus according to claim 11, further comprising: When the logical cluster to which the data block to be written belongs is registered in the address conversion table, the physical block number corresponding to the logical cluster number to which the data block registered in the address conversion table belongs is used. And access
The access is performed using a logical block number of the data block as a physical block number when a logical cluster to which the data block to be written belongs is not registered in the address conversion table. 11. A data writing method in the disk array control device according to 11.   12. The method of writing data in a disk array controller according to claim 11, wherein the total number of logical data blocks arranged in the indirect address cluster is limited within a range of the total number of entries in the address translation table. Determine whether there is a free physical cluster to be written next, or whether there is a free physical block in the current physical cluster to be written,
If it is determined that there is an indirect address data block written to a physical address on the disk different from the requested logical address in the write candidate logical block,
When it is determined that the indirect address data block exists in the write candidate logical block, the indirect address data block belonging to the physical cluster having the smallest number of effective blocks is selected,
12. The data writing method in the disk array control device according to claim 11, wherein the selected indirect address data block is written as the data block to be written.   When it is determined that there is no free physical cluster, or there is no free physical block in the current write target physical cluster, the write target block is registered in the address conversion table as the write target block. 17. The data writing method in the disk array control device according to claim 16, further comprising a step of writing to a physical block address corresponding to the data block to be written. If it is determined that the indirect address data block written to the physical address on the disk different from the requested logical address does not exist in the write candidate logical block, the address among the direct address data blocks of the write candidate Select the longest sequence of direct address data blocks that
17. The disk array control according to claim 16, further comprising a step of writing a column of the selected direct address data block when the column of the selected direct address data block satisfies a predetermined condition. Data writing method in apparatus.   When it is determined that the predetermined condition is not satisfied, the dirty rate of the cache memory that temporarily stores data to be written to the disk array and the use of the address conversion table for each direct address data block of the write candidate 19. The data writing method in the disk array control apparatus according to claim 18, further comprising a step of determining a writing method based on the rate.   The direct address data block as the write candidate is selected in order from a direct address data block belonging to a cache column having a high usage rate when it is determined that the predetermined condition is not satisfied. 20. A data writing method in the disk array control device according to 19.

Images