JPH06119123A - Disk array device - Google Patents

Abstract

PURPOSE:To prevent or evade a deadlock on the disk array device of multiple- rank and dual-port constitution. CONSTITUTION:First and second disk array control means 10-1 and 10-2 update data D1 of one disk device and a transaction 2 updates the parity P2 of the other disk and the parity P1 of the former disk device; when the condition where the transaction 2 or the data D2 of the latter disk device are updated are satisfied, the deadlock is entered if the update of the parity data is performed first, so a deadlock inhibiting means 60 inhibits the update of the parity from being performed first or releases the disk at the end of a parity read even if the update of the parity is performed first, thereby preventing or evading the deadlock.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a disk array device in which a plurality of disk devices are accessed in parallel to perform data input / output processing, and more particularly to a disk array device having a plurality of ranks and a dual port configuration. As external storage devices of computer systems, disk devices such as magnetic disk devices and optical disk devices, which have features such as non-volatility of recording, large capacity, and high speed of data transfer, are widely used. The requirements for the disk device are high-speed data transfer, high reliability, large capacity, and low price. A disk array device has been attracting attention as one that meets these requirements. A disk array device is a device in which a plurality of small disk devices are arranged, data is distributed and recorded in these devices, and accessed in parallel.

When data is transferred to a plurality of disk devices in parallel by a disk array device, it is possible to transfer data at a high speed that is twice as many as the number of disks as compared with the case of a single disk device. In addition to data, by recording redundant information such as parity data, it is possible to detect and correct data errors caused by disk device failure, etc., and duplicate the contents of the disk device. The same high reliability as the method of recording can be realized at a lower price than the duplication.

[0003]

2. Description of the Related Art Conventionally, David A. Passon of the University of California, Berkeley.
et al., published a paper that evaluated a disk array device that can access a large amount of data to many disks at high speed and realize data redundancy in the event of a disk failure, by classifying the disk array devices from level 1 to level 5. Announced (ACM SIGMOD Conferance, Chicago, Illi
nois, June 1-3, 1988 P109-P116).

Levels 1 to 5 for classifying the disk array device proposed by David A. Patterson and others are RA
ID (Redundant Arrays of In
abbreviated to "expensive Disks" 1 to 5. The RAID 1 to 5 will be briefly described as follows. [RAID0] FIG. 16 shows a disk array device having no data redundancy, which is not included in the classification of Debit A. Patterson et al.
Call it D0.

In the RAID 0 disk array device, as shown by data A to I, the disk array control device 10 distributes the data to the disk devices 32-1 to 32-3 based on the input / output request from the host computer 18. There is no data redundancy in the event of a disk failure. [RAID1] The RAID 1 disk array device is shown in FIG.
As shown in FIG. 7, a mirror disk device 32-2 that stores copies A ′ to D ′ of the data A to D stored in the disk device 32-1 is provided. RAID 1 is widely used because it has low utilization efficiency of a disk device but has redundancy and can be realized by simple control. [RAID2] A RAID2 disk array device strips data in units of bits or bytes,
Read and write in parallel to each disk unit. The striped data is physically recorded in the same sector in all disk devices. A Hamming code generated from the data is used as the error correction code. In addition to the data disk device, a disk device for recording the humming code is provided, and the failed disk device is specified from the humming code to restore the data.

By providing the redundancy by the hamming code as described above, correct data can be secured even if the disk device fails, but it has not been put into practical use because the utilization efficiency of the disk device is poor. [RAID3] The RAID3 disk array device has the configuration shown in FIG. That is, as shown in FIG. 19, for example, the data a, b, and c are divided into data a1 to a3, b1 to b3, and c1 to c3 in units of bits or sectors, and the parity P1 is calculated from the data a1 to a3 to obtain the data. b1
To b3, the parity P2 is calculated, the parity c3 is calculated from the data c1 to c3, and the disk device 32- of FIG.
1 to 32-4 are simultaneously accessed and written in parallel.

In RAID 3, data redundancy is maintained by parity. Further, the data writing time can be shortened by the parallel processing of the divided data. However, all the disk devices 3 can be accessed with a single write or read access.
A parallel seek operation of 2-1 to 32-4 is required.
For this reason, it is effective when handling a large amount of data continuously, but in the case of transaction processing in which a small amount of data is randomly accessed, the high speed of data transfer cannot be utilized and efficiency decreases. [RAID4] As shown in FIG. 20, the RAID4 disk array device divides one data into sector units and writes the data in the same disk device. For example, the disk device 32
Looking at -1, the data a is divided and written into sector data a1 to a4. The parity is stored in the fixedly determined disk device 32-4. Here, data a1,
Parity P1 is calculated from b1 and c1, and data a2 and
Parity P2 is calculated from b2 and c2, and data a3 and
Parity P3 is calculated from b3 and c3, and data a4 and
Parity P4 is calculated from b4 and c4.

Data reading is performed by the disk devices 32-1.
32-3 can be read in parallel. Data a ~
To read b, taking the data a as an example, the disk device 3
2-1 sectors 0 to 3 are accessed to access sector data a1
To a4 are sequentially read and combined. For data writing, after reading the data and parity before writing and calculating and writing new parity,
A total of 4 accesses are required. For example, when updating (rewriting) the sector data a1 of the disk device 32-1, the old data (a1) _old of the update location and the corresponding old parity (P1) _old of the disk device 32-4 are read and new The operation of writing for a new parity (P1) _new that is consistent with the data (a1) _new is required in addition to writing data for updating.

Further, since the access is always made to the parity disk device 32-4 at the time of writing, it is impossible to simultaneously execute the writing to a plurality of disk devices. For example, even if the data a1 of the disk device 32-1 and the data b2 of the disk device 32-2 are simultaneously written, the parities P1 and P2 from the same disk device 32-4 are written.
Cannot be written at the same time because it must be read and written after calculation.

Although the definition of RAID4 has been made in this way, there are few merits and there are few movements for practical use at present. [RAID5] The RAID5 disk array device enables parallel reading and writing by not fixing the disk device for parity. That is, as shown in FIG.
The disk device in which the parity is placed is different for each sector. Here, from data a1, b1, c1 to parity P
1 is calculated, and the parity P is calculated from the data a2, b2, d2.
2 is calculated, and the parity P is calculated from the data a3, c3, d3.
3 is calculated, and the parity P is calculated from the data b4, c4, d4.
4 has been calculated.

For parallel reading and writing, for example, the disk device 3 is used.
2-1 sector 0 data a1 and disk device 32-2
Since the data b2 of sector 1 is located in the disk devices 32-4 and 32-3 having different parities P1 and P2, the data b2 does not overlap and can be read and written at the same time. Note that the overhead that requires a total of 4 accesses at the time of writing is RAI.
Same as D4.

As described above, RAID 5 is capable of asynchronously accessing a plurality of disk devices to perform read / write, and therefore is suitable for transaction processing for randomly accessing a small amount of data. RAID3 to RAI
In the disk array device as shown in D5, a combination of disk devices related to the generation of redundant information will be called a rank. For example, when there are k disk devices for recording data and m disk devices for recording redundant information related to data, (k + m) disk devices are collectively ranked. Processing performance can be further improved by connecting a plurality of ranks of disk arrays.

[0013]

However, in a disk array device of a plurality of ranks, if the access path to each disk device is dual ported, there is a problem that a deadlock occurs during the disk writing process. It was The reason for this will be described below together with the RAID level 5 write processing. In RAID level 5, the exclusive OR of the data in each disk device is obtained and stored in the disk device as parity (error detection code) as in the following expression (1). Data a (+) Data b (+). ． ． = Parity P1 (1) However, (+) indicates an exclusive OR, and the storage locations of the data and the parity are such that the parities P1 to P4 are distributed to the respective disk devices 32-1 to 32-4 as shown in FIG. , Concentration of access to one disk device due to parity read / write operation (parity update operation) is eliminated. When data of RAID level 5 is read, the data in the disk device is not rewritten, so that the consistency of parity is maintained, but at the time of writing, the parity also needs to be changed according to the data.

When data is updated by rewriting one old data in a certain disk device to new data, in order to obtain the consistency of parity, the following formula (2) is calculated and the parity is updated. The parity consistency of the entire data can be maintained. Old data (+) Old parity (+) New data = New parity (2) As can be seen from the equation (2), in the data writing process,
It is necessary to first read the old data and old parity in the disk device. Further, since the writing is performed at the same place as the read-out place, when the writing operation of the disk device is started, it always takes a time for one rotation of the disk.

Further, the disk device in which the parity is written is
In order to generate the new parity from the equation (2), it is necessary to wait until the old data is read and transferred from the disk device which writes the data. FIG. 22 is a flowchart showing the processing operation of the RAID 5 level disk array device, and shows in parallel the processing operation of the data storage disk device and the parity storage disk device that cooperate with the processing of the disk array control device.

Next, the reason why deadlock occurs will be described. First, the processing when the host makes two write requests (hereinafter referred to as “transactions”) in the single-port disk array device will be described.
FIG. 23 shows a single-port disk array device in which the disk array 46 has a 4-rank structure and only one disk array control device 10 is provided.

In FIG. 23, 32-1 to 32-20 are disk devices, 34-1 to 34-5 are interfaces,
48-1 to 48-4 are ranks indicating array units.
Two disk devices 32-4, 32-
In the usage state of 17, the transaction 1 is generated as an update instruction for updating the data D1 of the disk device 32-9 and the parity P1 of the disk device 32-7 in the disk array control device 10, and then the disk device 32. -7
Transaction 2 as the update instruction for updating the data D2 of the disk device 32-9 and the parity P2 of the disk device 32-9.
However, it is assumed that they are successively generated with almost no time.

Here, the disk devices 32-4, 32-17
Since the interfaces 34-2 and 34-4 are in use, queues for each instruction of transactions 1 and 2 can be made until they are released. Disk device 32-4, 32-1
When 7 is released, the interfaces 34-2, 34-4
In the transaction 1 via the disk device 32-7,
The usage right of 32-9 is acquired.

The D1 update instruction of transaction 1 reads the old data D1 of the disk device 32-9, writes new data D1 next, and releases the disk device 32-9. Further, the P1 update command of transaction 1 reads the old parity P1 from the disk device 32-7, and if the old data D1 is read in this state, the new parity P1 is generated from the equation (2). Parity P
After writing 1, the disk device 32-7 is released.
Disk device 32-7, 32 according to transaction 1
The -9 accesses are performed concurrently in parallel.

After the processing of transaction 1 is completed, the D2 update instruction and the P2 update instruction of transaction 2 are similarly executed. Thus, in the single port configuration,
Since the transaction 2 generated later does not use the disk device first, the deadlock cannot occur. FIG. 24 shows a disk array device having a dual port configuration. Two disk array control devices 10-1 and 10 are provided.
-2 is provided, and interfaces 34-1 to 34-5,
Two access paths are formed by 36-1 to 36-5, and in principle, double the processing capacity is provided as compared with the single port configuration.

When access from the disk array controllers 10-1 and 10-2 to the same disk device is duplicated, the one who has acquired the usage right first occupies the other, and the other is occupied until the disk is released by the end of access. You will have to wait. The deadlock in the dual port configuration shown in FIG. 24 will be described below. Now, as shown in FIG.
The disk array control device 10-1 is the disk device 32-
4 and the disk control device 10-2 is using the disk device 32-17, the disk array control device 10- is continuously operated with almost no time.
It is assumed that transaction 1 has occurred in 1 and transaction 2 has occurred in the disk array control device 10-2.

The transaction 1 generated earlier is a D1 update instruction for updating the data D1 of the disk device 32-9 and P1 for updating the parity P1 of the disk device 32-7.
It has an update command and is queued. The transaction 2 generated later is a D2 update command for updating the data D2 of the disk device 32-7 and the disk device 32-9.
Has a P2 update instruction to update the parity P2 of the same, and is also queued.

Here, the interface 3 from the disk array control device 10-1 to the disk device 32-7.
Since 4-2 is vacant, the right to use the disk device 32-7 is acquired by the P1 update command, and the parity P
Can lead one. In addition, the disk device 32-
Since the interface 34-4 from the disk array control device 10-2 for 9 is also free, the usage right of the disk device 32-9 can be acquired by the P2 update command and the parity P2 can be immediately read.

When the parities P1 and P2 are first read in this way, the transactions 1 and 2 respectively take the occupied states of the disk devices 32-7 and 32-9 in order to write the newly generated parity. This is maintained and the state shown in FIG. 25 is obtained. Next, the transactions 1 and 2 are the old data D before being updated by the D1 update command and the D2 update command.
1 and D2 are read and an attempt is made to generate new parity.
However, even if an attempt is made to access the disk device 32-9 by the D1 update instruction of transaction 1, it cannot be accessed because it is already occupied by the P2 update instruction of transaction 2. At the same time, if an attempt is made to access the disk device 32-7 by the D2 update command of transaction 2,
It is already occupied by the P1 update instruction of transaction 1 and cannot be accessed in the same manner.

That is, the transactions 1 and 2 occupy their respective disk devices for parity update, while waiting for their ends to end by issuing an acquisition request to use the disk devices occupied by the other party. , A deadlock occurs in which neither disk unit is released and the process stops. In other words, a deadlock occurs in a cyclic standby state in an uninterruptible disk device that can perform only one process at a time. An object of the present invention is to provide a multi-rank, dual port, individual access to a disk device within a rank, and avoid deadlock in a disk array configuration in which parity is distributed. Disk array device.

Another object of the present invention is to provide a disk array device which judges the possibility of establishment of a specific deadlock from the contents of consecutively generated transactions and processes it. Another object of the present invention is to provide a disk array device in which a transaction processing procedure in which deadlock cannot occur is set.

Another object of the present invention is to provide a disk array device adapted to change the processing procedure of the subsequent transaction so as to avoid the deadlock when determining the possibility of establishing the deadlock. Another object of the present invention is to provide a disk array device which avoids deadlock by processing transactions in sequence.

[0028]

FIG. 1 is a diagram for explaining the principle of the present invention. First, according to the present invention, a disk having a disk array 46 having a plurality of sets of a plurality of disk devices 32 storing a plurality of data to be processed in parallel and a parity as redundant information of the data as one array unit (rank). Targets array devices.

The parity stored in each disk device 32 of the disk array 24 has a data storage format corresponding to the RAID 5 level and a parity generation format stored in a different disk device each time the data storage position changes. Further, at least two disk array control means 10-
1 and 10-2 are provided, and a dual port configuration having at least two access paths for individually accessing the disk device 32 from different paths to read / write data or parity is provided.

RA having such a plurality of ranks and dual ports, and further distributing the parity storage disk devices
In the present invention, a disk array control means 10-1, 10-2 for a disk array device corresponding to the ID5 level is provided.
Thus, deadlock suppressing means 34 for suppressing deadlock when transactions 1 and 2 as two processing requests for simultaneously accessing two disk devices belonging to the same array unit exist before and after is provided. To do.

Here, the deadlock has transactions 1 and 2 generated before and after the disk array processing means 10-1 and 10-2, and the transaction 1 updates the data D1 of one of the disk devices and the transaction 2 occurs. The parity P2 of the other disk is updated, and the transaction 2 has the parity P2 of one disk device.
This may occur when the condition that the transaction 2 or the data D2 of the other disk device is to be updated while the 1 is being updated is satisfied.

That is, if the transactions 1 and 2 both update the parity data first with the deadlock occurrence condition established, deadlock will occur. Therefore, the deadlock can be avoided by not starting the parity update process or by releasing the disk at the end of parity read even if the parity update is performed first.
The processing by the deadlock suppressing means 34 is described in the following case 1
There is 5. [Case 1] Old data D1, D2 and old parity P
The reading of 1 and P2 and the writing of the new parity are performed separately. That is, the disk device for parity is released when the reading of the old parities P1 and P2 is completed, and the data D1 and D2 can be updated by the other transaction. Once the data D1 and D2 have been updated, the old data D
Since 1 and D2 have already been read, new parity P1 and P2 may be generated, the disk device may be acquired again, and the new parity may be written. [Case 2] A procedure in which the parity storage disk cannot be accessed unless the data storage disk device and the parity storage disk device are simultaneously accessed in parallel and the data storage disk device is always accessed. Determine. [Case 3] The access progress status of the disk device for data is referred to when the previous parity P1 is read in the preceding transaction 1, and the disk device is occupied and released according to the following cases.

When the data storage disk device is accessing, for example, seeking the old data D1 and D2, transferring the buffer, transferring the old data to the buffer, etc., the old data is read by the data update after waiting for a while, and the new parity is set. Since the disk device can be released by generating and writing it, the disk device is continuously occupied until the parity update is completed.

If the disk device for data cannot be acquired and is in the queue, the condition for establishing the deadlock is judged. If the condition is satisfied, the transaction 2 generated later is currently occupied. The disk device for parity is released, and the data update by the transaction 1 that occurred earlier is processed first. [Case 4] A deadlock satisfaction condition is determined before accessing the disk device for parity in the subsequent transaction 2, and if it is satisfied, the access progress of the data disk device of the preceding transaction 1 is referred to Until the access to the parity disk device is started, the access to the parity disk device by the subsequent transaction 1 is prohibited. [Case 5] Before allocating all transactions to the disk array control means 10-1 and 10-2, the transactions are stored in the order of occurrence in the first-in first-out queue, and are sequentially stored in the disk array control means 10-1 and 10-2. Deliver and execute. Therefore, an access request to a disk device for a transaction that occurs later is not processed earlier than an access request to the disk device for a transaction that occurs earlier, and deadlock cannot occur.

[0035]

In the disk array control device of the present invention having such a configuration, so-called RAID 5 is possible in which each disk device of the disk array can be individually read and written and the parity is changed every time the position of the stored data is different. For a level-equivalent configuration, when multiple ranks and dual-port configurations are used, deadlocks that occur under specific conditions are predicted and deadlocks are prevented, so deadlocks are reliably prevented. it can.

Even when deadlock occurs, by forcibly releasing the disk, it is possible to avoid the deadlock or quickly recover from the deadlock state.

[0037]

【Example】

<Table of contents> 1. Hardware configuration of the present invention 2. Processing function of the present invention 3. 3. First embodiment of deadlock suppression processing of the present invention Second Embodiment of Deadlock Suppression Processing of the Present Invention 5. Third embodiment of deadlock suppression processing according to the present invention 4. Fourth embodiment of deadlock suppression processing according to the present invention Fifth embodiment of deadlock suppression processing of the present invention Hardware Configuration of the Present Invention FIG. 2 is a configuration diagram of an embodiment showing the hardware configuration of the disk array device of the present invention.

In FIG. 2, the disk array controller 1
0 is provided with two systems of disk array control units. That is, two MPUs 12-1 and 12-2 are provided as control means in the disk array controller 10 and M
Taking the PU 12-1 side as an example, the ROM 2 fixedly storing the host interface 16-1 for communicating with the host computer 18 on the internal bus 14-1, and the program and the like.
0-1, R used as control store or buffer
AM 22-1, cache memory 26-1 connected via cache control unit 24-1, data transfer buffer 2
8-1 is connected.

Similarly, the MPU 12-2 also has a host interface 16-2, a ROM 20-2, a RAM 22-2,
Cache memory 2 via cache control unit 24-2
6-2 and the data transfer buffer 28-2 are connected. On the other hand, as the disk array 46, in this embodiment, it is composed of 20 disk devices 32-1 to 32-20, and 5 disks form one rank, and rank 48-1.
It has 4 ranks of ~ 48-4. In addition, the disk device 32
-1 to 32-20 have a dual port configuration that can be accessed from two different systems.

Each disk device 32 of the disk array 46
-1 to 32-20 are device adapters 30-11 to 30-15 and 30-21 to 30-25 functioning as interface control units provided in the disk array control device 10.
Internal bus 14- of MPU 12-1, 12-2 via
1 and 14-2, respectively. Further, a dual port RAM 56 that functions as a communication control memory is connected between the internal buses 14-1 and 14-2. 2. Processing Functions of the Present Invention FIG. 3 is a functional block diagram showing the processing functions of the disk array device according to the present invention.

In FIG. 3, first, two systems of disk array controllers 10-1 and 10-2 are individually provided so that control signals and data can be exchanged with the host computer 18, respectively. The disk array control device 10-1 is provided with a control unit 38-1 realized by program control of the MPU, and in the control unit 38-1, a control storage 40-1, a data processing unit 44-1 and a parity processing unit. The function as 45-1 is provided.

The data processing unit 44-1 determines the logical I based on the result of decoding the write command from the host computer 18.
A process of rewriting stored data in a specific disk device designated by D and a sector address is executed. Further, the parity processing unit 45-1 performs a parity writing process in accordance with the data rewriting by the data processing unit 44-1. Here, since the data storage format and the parity generation format in the disk array device of the present invention are equivalent to the RAID5 level shown in FIG. 16, the data processing unit 44-1 is old in order to generate new parity. Data read processing is necessary, and the parity processing unit 45-1 also needs old data read processing.

In more detail, the control unit 38-
When a host command is given as a processing request from the host computer 18 to 1, the controller 38-1 creates a data update instruction and a parity update instruction and issues them to the specified lower disk device. Here, the data update instruction is data update instruction = (data read instruction) + (data write instruction), and the parity update instruction is parity update instruction = (parity read instruction) + (parity generation instruction) + (parity Write command). The data processing unit 44-1 issues the data read instruction in the data update instruction as one package instruction to the target disk device to perform data read and data write access.

Further, the parity processing unit 45 first issues a write command to the target disk device, and when the old parity can be read by this, the data processing unit 44-1 confirms the reading of the old data to generate the parity. Part 50-1
When the new parity is generated, a parity write command is issued to write the new parity.

The parity generation unit 50-1 exclusively uses the old data read from the write target disk, the old parity read from the same sector position of another disk belonging to the same rank, and the new data given from the host computer 18. Generate a new parity from the logical sum. Further, the disk control device 10-1 is provided with a buffer unit 52-1 and an interface control unit 54-1. The buffer unit 52-1 stores new data, old data, and old parity when writing data. The interface control unit 54-1 controls the disk devices 32-1 to 32-in the disk array.
access n.

Such a disk array controller 10-
The configuration of 1 is the same on the side of the disk array control device 10-2. Further, the disk array controller 10-
In order to enable exchange of control information between 1 and 10-2, a communication control storage 56 using a dual port RAM or the like is provided, and the control unit 3 for the communication control storage 56 is provided.
By reading and writing by 8-1 and 38-2, necessary information can be exchanged with each other.

In the disk array 46 in the functional block diagram of FIG. 3, disk devices 32-1, 32-2, ...
Only one rank of 32-n is shown, and among them, transactions 1 and 2 as processing requests for writing two data to the disk devices 32-1 and 32-2 are disk array control devices 10-1 and 10-. Data D corresponding to the deadlock occurrence condition that occurs when it occurs for each of 2
1, D2 and parities P1 and P2 are stored.

Disk array controller 10-1, 10-
The deadlock management tables 42-1 and 42-2 are provided in the control storages 40-1 and 40-2 provided in FIG.
FIG. 4 is an explanatory diagram of a deadlock management table used in the present invention, in which information on transaction numbers, parity storage disks, and data storage disks is written. Further, the contents of the parity storage disk and the data storage disk are divided into two, a disk number and access information.

FIG. 5 shows an operating state immediately before a deadlock occurs in the disk array device of the present invention.
The state is the same as the description of the deadlock in FIG. 19, except that information can be exchanged between the disk array control devices 10-1 and 10-2. That is, in the state of FIG. 5, the preceding disk array control device 10
Disks 32-4, 32-17 are in use by two transactions at -1, 10-2. If the disks 32-4 and 32-17 are for parity and the new parity is being written, the data disks are, for example, 32-1 and 32-20, and it is assumed that the data writing has been completed respectively. To do.

In this state, transaction 1 and transaction 2 occur, and transaction 1 is the disk 3
2-9 data writing, so transaction 1 queues the D1 update command for disk 32-9 and the P1 update command for disk unit 32-7. In the disk array control device 10-2, a D2 update command for the disk device 32-7 and a P2 update command for the disk device 32-9 by the transaction 2 which occur with almost no time in the transaction 1 are queued.

In the operation state as shown in FIG. 5, the deadlock management table used in the present invention has the registered contents as shown in FIG. Therefore, by referring to the deadlock management table, it is possible to determine whether the deadlock occurrence condition is satisfied or know the disk access status. 3. First Embodiment of Deadlock Suppression Processing of the Present Invention FIG. 6 is a time chart showing a first embodiment for deadlock suppression of the present invention.

First, the deadlock is caused by the disk device 32-
1, 32-2 data D1, D2 and parity P1,
Depending on the storage status of P2, consecutive transactions 1,
2 occurs in the disk array controllers 10-1 and 10-2, and occurs when the parity P1 and P2 are updated. Therefore, in the first embodiment, as shown in FIG. 6, the parity P from each of the disk devices 32-1 and 32-2.
When P1, P2 are read first, the parity read is completed, the disk devices 32-1, 32-2 are released, and the data D is sent to the other side.
Access (read / write) of 1 and D2 is performed.

When the data D1 and D2 have been updated by reading and writing and the disk devices 32-1 and 32-2 are released, the disk device is acquired again and the calculated new parity is written. FIG. 7 is a flow chart showing the processing of the first embodiment for suppressing deadlock in the present invention. The processing operation of the disk array controller shown in FIG. 7 is performed by the disk array controllers 10-1, 10- of FIG.
2 will be executed in parallel.

First, the disk array control device receives a write command from the host in step S101. Taking the disk array control device 10-1 as an example, the disk array control device writes the write command to the area of the data D1 of the disk device 32-2. To receive. Then, in step S102, a data update command is issued to the disk device 32-2 designated by the logical ID of the write command. Specifically, in the process of issuing the data update command, first, an access request is issued to the disk device 32-2, and if it is a busy response, it is placed in a queue.
When the ready response is obtained, the access right is acquired, the connection is made through the interface, and then the data update command is transmitted.

Further, since the disk device 32-2 for data writing and the data writing sector are specified in step S103, the disk 32-1 storing the parity data at this sector position is specified, Disk device 32
Issue a parity update instruction to -1. As with the data update instruction, this parity update instruction also issues an access request to the disk device 32-1 and puts it in a queue if it is a busy response, and acquires an access right when a ready response is obtained. To send a parity update command.

The processing of the data storage disk device activated by the data update command is performed in steps S201 to S206.
As shown in. That is, in step S201, a seek for positioning the head at the beginning of the designated write block is executed, and in step S202, when the read is completed due to the completion of seek, the old data is read out in the buffer section of the disk array controller in step S203. Read and transfer.

Subsequently, when one revolution is waited in step S204,
Since writing becomes possible in step S205, new data is written to the disk in step S206, the series of processing is terminated, and the disk device in which the data has been written is released. On the other hand, in the parity storage disk device activated by the parity update command, the seek process is started in step S301, and if the head can be positioned at the head of the read block in step S302, it becomes readable, and the parity data is read in step S303. Read and transfer to the buffer on the disk array controller side. Subsequently, in step S304, the disk is rotated once, and then in step S305, it is checked whether or not a new parity is generated.

On the disk array controller side, new parity is generated and transferred when the reading of the old data and the old parity is completed. Therefore, step S30
When the head is positioned at the head of the block and is writable by the disk rotation in step S304, the parity data is written to the disk in step S307 and the parity storage disk device is opened.

The basic processing operation of the data storage disk device and the parity storage disk device is the same as the conventional one, but the parity storage is performed in the first embodiment by the processing of the disk array control device. The disk device is forcibly released after the parity data is read out and transferred to the buffer in step S303 of the disk device.

For this reason, in step S103 of the disk array controller, it is monitored whether or not the reading of the old parity is completed after the parity update command is issued, and when the old parity is read and transferred to the buffer, the step is executed. In step S104, the disk device on the parity side is forcibly released. With such forced release of the disk device on the parity side, at this time, since the data update command for transaction 2 is queued in the disk array control device 10-2 side by side at the same time, the access right is granted after the disk is released. Is obtained, and the access process by the data update command of transaction 2 is performed.

When the data update process on the transaction 2 side is completed, the disk device is released, so the disk array control device monitors in step S107 whether or not the disk device on the parity side can be accessed, and access is possible. If so, the access on the parity side is restarted in step S108, and the processing from step S304 in the parity disk device is restarted.

In the meantime, in the disk array controller, the completion of reading the data and the parity is determined in step S105, and therefore the parity generation process is performed in step S106, and the disk access on the parity side is performed in step S108. When restarting, new parity will be sent from the disk array control device, and when writing becomes possible from step S305 to step S306,
Immediately in step S307, the new parity can be written.

The disk array control device executes step S10.
After the access to the disk device on the parity side is restarted in step 8, the completion of writing the data and the parity is monitored in step S109. When the writing of both is completed, the host computer is notified of the completion of writing and a series of processing is performed. finish. 4. Second Embodiment of Deadlock Suppression Processing of the Present Invention FIG. 8 is a time chart showing a second embodiment of deadlock suppression processing of the present invention. In the second embodiment, when a transaction occurs regardless of whether or not the deadlock occurrence condition is satisfied, the access of the disk device for data storage is started and then the disk device for parity storage is started. The feature is that the access is started. For example, when transaction 1 occurs, access to the disk device 32-1 for storing data is started and then access 32-1 to the disk device for storing parity is started. The same applies to the transaction 2.

Specifically, taking transaction 1 as an example, processing is performed so that access to the disk device 32-1 for storing parity is not permitted until it is determined that the disk device 32-2 for storing data can be accessed. To do. FIG. 9 is a flow chart showing a second embodiment of the deadlock suppression processing of the present invention.

In the disk array control device, when the host command is received in step S101, it is determined in step S102 whether or not the disk device on the data side can be accessed. When the disk device on the data side obtains the ready response and acquires the access point, the process proceeds to step S103, and a data update command is issued to the disk device for storing data. Further, the fact that the access is started is written in the corresponding access state of the deadlock management table shown in FIG.

Next, in step S104, it is checked whether the disk device on the parity side is accessible, and if the access right is acquired by the ready response, the process proceeds to step S105, and a parity update command is issued. Subsequent processing waits for the completion of reading the old data and the old parity in step S106, generates the parity in step S107, finally confirms the writing of the data and parity, and ends the series of processing. Further, the processing in the data storage disk device and the parity storage disk device is the same as that of the first embodiment except that the disk release is not performed in the middle, and when data writing and parity writing are completed, FIG. Rewrite the access state of the deadlock management table shown to written. 5. Third Embodiment of Deadlock Suppression Processing According to the Present Invention FIG. 10 is a flowchart showing a third embodiment of deadlock suppression processing according to the present invention. In the third embodiment, when the parity data P1 is read from the disk device 32-1 in the transaction 1 that occurred earlier, the progress status of access to the data storage disk device 32-2 in the same transaction 1 is referred to. , If it is in use, wait a while and then the data storage disk unit 3
Since the old data is read from 2-2 and the new parity can be calculated, the occupied state of the disk device 32-1 for storing the parity is maintained, the parity write is continued, and then the disk device 32-1 is released.

On the other hand, as shown in the time chart of FIG. 11, the data storage disk device 3 of transaction 1
When the disk device 32-2 is occupied by the subsequent transaction 2 in the read state of the parity P1 of 2-1 and the update instruction of the data D1 is in the queue, it is necessary to establish the deadlock. The occupation of the disk device 32-2 by the transaction 2 subsequent to the above is forcibly released, and the data update access of the preceding transaction 1 is terminated first.

FIG. 12 is a flow chart showing a third embodiment of the deadlock suppressing process according to the present invention. 12
The disk array controller shown in (1) is a process on the side that executes the transaction 1 that occurred earlier. First, when a command is received in step S101, a data update command is issued in step 102 to access the data storage disk device for data update.

Then, in step S103, the transaction number and the disk number of the parity disk device are notified and registered in the deadlock management table of the disk array control device of the other party which is processing the self and the subsequent transaction 2. To do. By the notification of the transaction number and the disk number, it is possible to determine whether or not the deadlock occurrence condition is satisfied from the deadlock management table.

Next, in step S104, a parity update instruction is issued to the parity storage disk device. Subsequently, the completion of reading the parity is monitored in step S105. When the old parity is read and transferred to the buffer, the process proceeds to step S106, and the access progress of the data storage disk device is referred to by the deadlock management table. .

At this time, based on the data update command issued in step S102, the data storage disk device is accessing during the seek in step S201, the readable state in step S202, and the buffer transfer of the read data in step S203. The old data can be obtained on the disk array controller side by waiting a short time.

Therefore, the process proceeds to step S110 while maintaining the control state of the parity storage disk device, the read confirmation of the old data and the old parity is performed, and step S11 is performed.
In 1, the parity is generated and sent to the parity storage disk device, the parity is written in the processing of steps S304 to S307, and the parity storage disk device is released.

Therefore, when the parity storage disk device is released, the access right of the disk device that executes the data update instruction of the subsequent transaction in the queue is acquired, and the deadlock does not occur. On the other hand, if the data storage disk device of transaction 1 is in the queue due to the occupation of the disk device due to the execution of the parity update instruction of the subsequent transaction 2 in step S107, the deadlock management table is referenced in step S108. It is determined whether the deadlock condition is satisfied.

If the deadlock condition is satisfied at this time, the process advances to step S109 to execute the data update instruction in which the disk device for parity storage occupied by the subsequent transaction 2 is queued in transaction 1. In order to open it, Therefore, the data update command issued in step S102 that has been in the queue receives the access right upon receiving the disk release on the transaction 2 side, and performs the data update process by the data storage disk device. When the disk device is released after the transaction 1 data update processing is completed, the parity update of the transaction 2 forcibly released in step S109 is restarted, and the parity generation and the parity write of the transaction 2 are performed. . 6. Fourth Embodiment of Deadlock Suppression Processing According to the Present Invention FIG. 13 is a time chart showing a fourth embodiment of deadlock suppression processing according to the present invention. In the fourth embodiment, it is determined whether or not the deadlock condition is satisfied before accessing the parity storage disk device 32-2 in the transaction 2 that occurs later, and if it is satisfied,
It is characterized in that the parity access of the transaction 2 is not permitted until the data access of the transaction 1 is started with reference to the access progress status of the data storage disk device 32-1 of the transaction 1 which has occurred earlier.

That is, the third embodiment of FIGS. 110 to 12 is the processing of the preceding transaction 1 by the disk array controller, whereas the fourth embodiment of FIG. 13 is the disk for executing the transaction 2 generated later. The processing is performed by the array controller. FIG. 14 is a flow chart showing a fourth embodiment of the deadlock suppressing process according to the present invention. First, when a command from the host computer is received in step S101 and the subsequent transaction 2 occurs, a data update command is issued in step S102. Issue and start access to the data storage disk device.

Next, in step S103, it is determined whether or not the deadlock condition is satisfied by referring to the deadlock management table. If the deadlock condition is satisfied, the progress of data access of the preceding transaction 1 is executed in step S104. Refer to the situation. At this time, if the data access of the preceding transaction 1 is not started and the data update instruction is in the queue, it waits until the data access is started.

When the data access is started in the preceding transaction 1, the parity access in the succeeding transaction 2 is permitted, and the parity update command is issued in step S106 to start the parity update processing by the parity storage disk device. After that, the reading of the old data and the old parity is confirmed in step S107, and the parity is generated in step S108.
When the writing of data and parity is confirmed in 9, the writing end and the disk number are written in the deadlock management table in step S110, and a series of processing is ended. 7. Fifth Embodiment of Deadlock Suppression Processing of the Present Invention FIG. 15 shows a FIFO command queue in a disk management table used in a fifth embodiment of deadlock avoidance processing of the present invention.

That is, according to the fifth embodiment of the present invention, before allocating the requests for all transactions to the disk devices, once for each access request to all the disk devices, as shown in FIG. It is characterized in that it is arranged in a command queue for various FIFOs.
In FIG. 15, a FIFO queue stores a data update instruction and a parity update instruction that realize the requests of transaction 1 and transaction 2, and the data update instruction is two access instructions of a data read instruction and a data write instruction. Therefore, this is stored in order. Further, since the parity update instruction is the three access instructions of the parity read instruction, the parity generation instruction, and the parity write instruction, these are also arranged in order.

Each access command entered in the FIFO command queue of FIG. 15 is fetched in order from the bottom to access the corresponding disk device, and the data update command and parity for the same disk device in transactions 1 and 2 are read. Since update instructions do not overlap, deadlock can be reliably prevented. In the above embodiment, 5 ranks are composed of 5 disk devices,
This is an example in which this is a 4-rank configuration and further a dual-port configuration in which two access buses are used for each disk device, but the present invention is not limited to this.
The number of disk devices, the number of ranks, and the number of access ports per rank can be appropriately determined as needed.

[0080]

As described above, according to the present invention, in a disk array device having a plurality of ranks and a dual port configuration, a deadlock that occurs when parity update and data update from different transactions for the same disk overlap. It is possible to prevent or avoid the above and appropriately execute the input / output processing, and it is possible to improve the throughput of the disk array devices operating in parallel by the dual port.

[Brief description of drawings]

FIG. 1 is an explanatory view of the principle of the present invention.

FIG. 2 is a configuration diagram of an embodiment showing a hardware configuration of the present invention.

FIG. 3 is a functional block diagram showing the processing functions of FIG.

FIG. 4 is an explanatory diagram of a deadlock management table used in the present invention.

FIG. 5 is an explanatory diagram showing a transaction occurrence state in which a deadlock occurs.

FIG. 6 is a time chart showing a first embodiment of deadlock suppression processing of the present invention.

FIG. 7 is a flowchart showing a first embodiment of deadlock suppression processing of the present invention.

FIG. 8 is a time chart showing a second embodiment of deadlock suppression processing of the present invention.

FIG. 9 is a flowchart showing a second embodiment of the deadlock suppression processing of the present invention.

FIG. 10 is a time chart showing a third embodiment of deadlock suppression processing of the present invention.

FIG. 11 is a time chart showing another deadlock suppressing process in the third embodiment.

FIG. 12 is a flowchart showing a third embodiment of deadlock suppression processing of the present invention.

FIG. 13 is a time chart showing a fourth embodiment of deadlock suppression processing of the present invention.

FIG. 14 is a flowchart showing a fourth embodiment of deadlock suppression processing of the present invention.

FIG. 15 is an explanatory diagram of control storage used in the fifth embodiment of the deadlock suppression processing of the present invention.

FIG. 16 is an explanatory diagram of a RAID 0 disk array device.

FIG. 17 is an explanatory diagram of a RAID 1 disk array device.

FIG. 18 is an explanatory diagram of a RAID 3 disk array device.

FIG. 19 is an explanatory diagram of data division in RAID3.

FIG. 20 is an explanatory diagram of a disk array device using RAID4.

FIG. 21 is an explanatory diagram of a disk array device based on RAID5.

FIG. 22 is a flowchart showing a data writing process of the disk array device by RAID5.

FIG. 23 is an explanatory diagram showing that deadlock does not occur in a single port configuration.

FIG. 24 is an explanatory diagram showing a state immediately before deadlock occurs in a dual port configuration.

FIG. 25 is an explanatory diagram of a state in which deadlock has occurred in the dual port configuration.

[Explanation of symbols]

10-1, 10-2: Disk array control device 12-1, 12-2: MPU 14-1, 14-2: Internal bus 15: System bus 16-1, 16-2: Host interface 18: Host computer ( Upper device) 20-1, 20-2: ROM 22-1, 22-2: Volatile memory (RAM) 24-1, 24-2: Cache control unit 26-1, 26-2: Cache memory 28-1 , 28-2: data transfer buffer 30-11 to 30-15, 30-21 to 30-25: device adapter 32-1 to 32-n: disk device 34-1 to 34-5, 36-1 to 36- 5: Interfaces 38-1, 38-2: Control Units 40-1, 40-2: Control Storage 42-1, 42-2: Deadlock Management Tables 44-1, 44-2: Data Processing Unit 45- , 45-2: Parity processing unit 46: Disk array 48-1 to 48-4: Rank 50-1, 50-2: Parity generation unit 52-1, 52-2: Buffer unit 54-1, 54-2: Interface control unit 56: control storage for communication 60: deadlock suppressing means

Claims (8)

[Claims] 1. A plurality of disk devices (32) storing a plurality of data processed in parallel and redundancy information of the data.
A plurality of disk arrays (46) as one array unit, and redundant information stored in each disk device of the disk array (46) is stored in a different disk device each time the data storage position changes, and a different path is stored. From at least two disk array control means (10) for individually accessing the disk device to read or write data or redundant information.
-1, 10-2) and a deadlock when there are two processing requests for simultaneously accessing two disk devices belonging to the same array unit by the disk array control means (10-1, 10-2). A disk array device comprising: deadlock suppressing means (34) for suppressing. 2. The disk array device according to claim 1, wherein the deadlock suppressing means (34) is generated before and after the two disk array processing means (10-1, 10-2). 1 and 2 processing requests, the first processing request updates data (D1) at a specific position of the first disk device, and the second processing request updates redundancy information (P2) at different positions of the first disk. , And the first processing request is the second
When the redundant information (P1) at a specific position of the disk device is updated and the second processing request is about to update the data (D2) at a different position of the second disk device, it is determined that there is a possibility of deadlock. Disk array device characterized by. 3. The disk array device according to claim 1, wherein the deadlock suppressing means (34) temporarily releases the disk device for parity after reading the redundant information (P1, P2) based on the processing request. , A data array (D1, D2) is read and written to update data before generating and writing redundant information. 4. The disk array device according to claim 1, wherein the deadlock suppressing means (34) first starts updating the data (D1, D2) based on the first and second processing requests, Redundancy information (P1, P
A disk array device characterized by permitting the start of updating of 2). 5. The disk array device according to claim 1, wherein the deadlock suppressing means (34) is configured to read the redundant information (P1) based on the first processing request generated earlier, at the time when the reading is completed. Referring to the progress status of the update of the data (D1) based on one processing request, the occupancy state of the disk device is maintained until the end during the data update, while the second processing request generated later in the data update wait state When the possibility of deadlock is determined from the relationship of the above, the disk device in use for updating the redundant information (P2) based on the second processing request generated later is released, and the first processing request generated earlier is released. Disk array device characterized by precedingly updating the data (D1). 6. The disk array device according to claim 1, wherein said deadlock suppressing means (34) dead before starting access to the disk device for updating redundant information by a second processing request generated later. When the possibility of locking is determined, the access progress status of the disk device for updating the data is referred to by the first processing request generated earlier, and the redundant information by the second processing request generated later is accessed until the access is started. A disk array device characterized by inhibiting access to a disk device to be updated. 7. The disk array device according to claim 1, wherein the deadlock suppressing means (34) sends a processing request to the first and second disk array control means (10-1,
10-2), the first or second disk array control means (10-1, 10-2) are arranged in a queue with a first-in first-out structure and are arranged in order from the processing request generated first.
A disk array device characterized by being executed by. 8. The disk array device according to claim 1, wherein said first and second disk array control means (10-
1, 10-2) uses parity data as redundant information and generates new parity data from exclusive OR of old data before update, old parity and new data to be updated. apparatus.