JP2008198221A - Automatic capacity expansion method for storage device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To determine a storage area of a storage device not uniquely but dynamically. <P>SOLUTION: This system monitors a logical block address of a read or write I/O accessed to the logical volume of the storage device from a host. The storage area of the logical volume is dynamically extended based on the acquired logical block address. The storage area of the logical volume is also reduced/expanded by an instruction to reduce/expand capacity of the logical volume to a volume server from a command part of the host. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a method for performing sequential capacity expansion while online in a storage device system.

  In recent years, the movement to store the storage areas of a plurality of host computers in one storage device has become popular. A typical storage device that promotes this movement is a disk array. The disk array enhances reliability by providing storage area redundancy by a plurality of magnetic disk storage devices within the apparatus, and provides a storage area having a necessary capacity to a plurality of hosts as a logical volume. One of the major advantages of using this disk array is that the capacity of the logical volume can be expanded.

  For example, when the computer uses up the logical volume provided by the disk array, the unused area inside the disk array is allocated as a logical volume cut out with an arbitrary length, and the computer combines this logical volume with the logical volume in use. The storage area can be expanded. This function is called online volume expansion because the volume can be expanded while online. With online volume expansion, the range of the logical volume storage area is expanded without stopping the application for data that increases every moment, so that the operation time of the application can be expanded. In addition, since it is not necessary to transfer data between volumes when transferring the volume capacity, storage management costs can be greatly reduced.

  Conventionally, in order to perform online volume expansion, a logical volume user must contact the logical volume provider to perform expansion. A small site that is used within a single company does not need to expand online volume at any time because there is no tight increase in data, but a large site that is used jointly by multiple companies such as a data center. Then, a sudden increase in data may occur from computers of multiple companies, and it is necessary to perform online volume expansion at any time. In addition, in order to increase the efficiency of use of storage areas such as disk arrays, one storage device can be used without waste

  In order to use the storage area of the storage device without waste, it is necessary to manage the storage area in units of a small capacity logical volume, and to expand the logical volume with a small capacity logical volume each time the expansion is performed. When online volume expansion is used under such circumstances, there are cases in which requests for online volume expansion are simultaneously generated from a plurality of users due to data that increases every moment, and the logical volume provider cannot be partitioned. In the worst case, there is a problem that online volume expansion cannot be performed and the processing of the computer is forced to stop.

  The logical volume user requests infinite storage capacity without managing the storage area. On the other hand, the logical volume provider requests to manage the storage area efficiently and provide it to the user more quickly without considering how the logical volume user uses the storage area.

  It is an object of the present invention to solve the demands of users and providers of storage resources by using a storage area management device, which is dynamically appropriate from a centrally managed storage area while the computer is online. The purpose is to expand the logical volume by allocating a storage area of capacity.

  The computer system of the present invention has a configuration in which one or more host computers, one or more disk storage devices, and a volume providing device are interconnected. A volume providing device that manages a plurality of disk storage devices provides a logical volume corresponding to each host computer from the plurality of disk storage devices. The host computer sends an I / O request to the logical volume, the volume provider reads the logical block address where the I / O request is read or written, and stores the logical block address accessed by the I / O request on the logical volume. When the area does not exist, the volume providing apparatus allocates a storage area from an unused magnetic disk storage device and dynamically expands the storage area of the logical volume. In addition, the logical volume is reduced by an arbitrary amount according to an application instruction.

  According to the present invention, by adding a storage area to a volume server in response to a request from the volume server, a host computer application can use a single volume whose capacity can be dynamically expanded and contracted.

<First embodiment>
Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 shows a configuration example of a computer system to which the present invention is applied. The computer system includes a plurality of hosts 1100, a volume server 1200, and a plurality of magnetic disk storage devices 1300. In this embodiment, there is one host 1100 and three magnetic disk storage devices 1300, but there may be one or more. In addition, the host 1100, the volume server 1200, and the magnetic disk storage device 1300 have a CPU, a memory, and the like.

  The host 1100 includes an application (App) 1110, an operating system (hereinafter referred to as OS) 1130, and a channel interface 1140. The application 1110 is application software such as a DB or a file system that performs read processing and write processing on a volume provided from the volume server 1200 that is a volume providing device. The operating system 1130 includes a volume device unit (rsd0) 1131 that receives an I / O request from the application 1110 and transfers it to a channel interface (I / F) 1140 and an instruction unit 1132. The command unit 1132 of the OS 1130 is software that manages logical partition information including the start block position and end block position of the storage device used by the OS 1130 and controls the operation of the volume server 1200.

  The logical partition information 4100 will be described with reference to FIG. 4 showing a logical volume which is a storage area provided from the volume server 1200. In this embodiment, in order to dynamically expand the volume, the start block position of the logical partition information handled by the OS 1130 is indicated by 0 LBA (4200), and the end block position is the capacity of the maximum storage device supported by the OS of the storage device or Indicates the storage capacity used by the user. If the maximum capacity is limited by the storage capacity used by the user, the storage area of the logical volume provided by the volume server 1200 is also limited by this end block position. Also, the logical partition information is stored in the storage device under the management of the OS 1130 from the start block position 0 LBA (4200) to the LBA for the area (4100) required for storing the logical partition information. Normally stored data is stored in a storage area (4400) other than logical partition information.

  A channel I / F 1140 shown in FIG. 1 is a device having a function of transferring I / O from the OS 1130 to an external device, and is a block device I / F such as a fiber channel I / F or a SCSI I / F.

  Commands for controlling between the host 1100 and the volume server 1200 and between the volume server 1200 and the disk 1300 are called I / O requests. The I / O request is a command of a block device protocol such as a SCSI protocol. FIG. 17 shows the command format of the I / O request. The I / O request has an operation code 17001, LUN (Logical Unit Number) 17002, LBA (Logical Block Address) 17003, and transfer data length 17004. The operation code 17001 is a number indicating read processing, write processing, and the like. The LUN 17002 is a unique number of the logical volume that is processed by the command. The LBA 17003 is a position where processing on the logical volume provided from the volume server 1200 is performed. The transfer data length 17004 is an amount processed at a time by this I / O request.

  The volume server 1200 includes a channel I / F 1210 on the host side, a volume providing unit 1230 that performs processing of I / O requests requested from the host 1100 and management of the magnetic disk storage device 1300, and an I requested by the volume providing unit 1230. It comprises a channel I / F 1220 for transferring the / O request to the magnetic disk storage device 1300. The host-side channel I / F 1200 is a block device I / F such as a fiber channel I / F or SCSI that can be connected to the host 1100. In this embodiment, the channel on the host 1100 side and the magnetic disk storage device 1300 side are separated, but the channel I / F 1210 on the host 1100 side and the channel I / F 1220 on the magnetic disk storage device 1300 side may be shared. .

  The volume server 1200 provides the host 1100 with a logical volume having infinite capacity. That is, when the volume server 1200 is viewed from the host 1100, the start value of the logical block number (LBA) that determines the volume range is determined to be 0, but the LBA that is the end value of the range may not be specified. . Further, when the LBA of the end value of the range is defined, the value becomes the maximum capacity of the volume provided from the volume server 1200.

  The volume providing unit 1230 in the volume server 1200 includes an I / O monitoring unit 1231, a capacity increasing / decreasing unit 1232, a segment management unit 1233, an I / O processing unit 1234, and a physical logical address management unit 1235. The segment management unit 1233 has a segment management table 2000 therein, and manages the storage area of the volume provided by the volume server 1200 using the segment management table 2000.

  The segment management table 2000 in FIG. 2 includes a disk number (Disk ID) 2100, a segment number (Segment Number) 2200, a start LBA (LBA (START)) 2300, a segment size (Size) 2400, and the usage status of the segment ( In-Use) 2500. The disk ID 2100 is a unique identification number of a disk determined by a block level protocol such as SCSI when the magnetic disk storage device 1300 is connected to the volume server 1200. A segment is a minimum unit of a storage area managed by the volume server 1200 divided in each magnetic disk storage device 1300, and a segment number 2200 is a unique identification number for managing the segment by the volume server 1200. . The LBA (START) 2300 defines the physical position where the segment storage area starts in the magnetic disk storage device 1300 with the disk ID 2100. The size of the segment indicates the size of the storage area starting from LBA (START) 2300 in the disk. The usage status of the segment indicates by two values whether the storage area of the segment is used by the volume server 1200 or not. The usage status is indicated by a value of 1 when in use, or 0 when not in use.

  Next, the physical logical address management unit 1235 performs physical logical address management between the logical volume provided by the volume server 1200 and the magnetic disk storage device 1300 in which data is stored. As a means for managing the magnetic disk storage device 1300 in which this logical volume and actual data exist, the physical logical address management unit 1235 has the physical logical management table 3000 shown in FIG. In the physical logical management table 3000, a logical unit number (LUN) 3100, a segment number (Segment Number) 3200, a LUN LBA (START) 3300, and a LUN LBA (END) 3400 are managed. The LUN 3100 is called a logical unit number and is a unique number of a logical volume provided from the volume server 1200 to the host 1100. The segment number 3200 is a unique number of the storage area managed by the segment management unit 1233. The logical volume provided by the LUN 3100 is composed of a plurality of segments, and the logical block number (LBA) of the logical volume is consecutive in the order of the young segment LBA to the old segment LBA in each LUN 3100 of the physical logical management table 3000. Are combined.

  For example, when LUN 0 (3500) in FIG. 3, segment number 1 (3510), segment number 2 (3520), and segment number 3 (3530) are combined in this order, and LUN 0 (4000) single logical volume as shown in FIG. To the host 1100. The LUN LBA (START) 3300 indicates the start position of the LBA used by the segment in the LUN. The LUN LBA (END) 3400 indicates the end position of the LBA used by the segment in the LUN. The I / O monitoring unit 1231 monitors LBAs in individual I / O requests accessed from the host 1100 to the magnetic disk storage device 1300. The I / O processing unit 1234 performs an actual read / write I / O process. The capacity increasing / decreasing unit 1232 performs a capacity increasing / decreasing process requested from the I / O monitoring unit 1231 or the command unit 1132 of the host 1100.

  In this embodiment, the volume server 1200 is arranged as an independent device, but the function of the volume server 1200 may be incorporated in the magnetic disk storage device 1300.

In this embodiment, volume expansion is performed on a segment basis. However, when the segment management unit 1233 and the physical logical address management unit 1235 are regarded as a file system whose capacity can be expanded, such as a Log Structured file system, the segment management table The logical unit with the LUN 3100 provided by the volume server 1200 is replaced by replacing the segment which is the smallest unit of the storage area in the 2000 and the physical logical management table 300 with the block which is the smallest unit of the storage area of the file system. It can be regarded as a volume. That is, even if the segment management unit 1233 and the physical logical address management unit 1235 are replaced with a file system, a logical volume that can be dynamically expanded in units of blocks can be provided.
(Description of operation)
Next, the operation of the volume server 1200 will be described with reference to FIG. After an I / O request is issued from the host 1100 to the volume server 1200 (step 5001), the volume server 1200 processes the received I / O request at the volume providing unit 1230 (step 5002). When the processing of the I / O request is completed, an I / O request processing completion notification is issued to the host 1100 (step 5003). The host receives the I / O completion notification (step 5004) and ends the process. At first glance, the volume providing unit 1230 of the volume server 1200 appears to operate in the same manner as the control unit of the conventional magnetic disk storage device 1300, but has a mechanism for infinitely expanding the volume. Hereinafter, an internal operation of the volume providing unit 1230 provided with this mechanism will be described.

  The volume providing unit 1230 operates in cooperation with the I / O monitoring unit 1231, the I / O processing unit 1234, the physical logical address management unit 1235, the capacity increase / decrease unit 1232, and the segment management unit 1233. A thick line in the enlarged view 1236 of the volume providing unit 1230 in FIG. 1 indicates a flow of I / O requests, and a thin line indicates a flow of control. First, an I / O request sent from the host-side I / F 1210 is processed by the I / O monitoring unit 1231, and then the I / O processing unit 1234 performs physical-to-logical conversion of the LBA of each I / O request, Read / write processing is performed on the magnetic disk storage device 1300. Each part will be described in accordance with this processing flow.

  The operation of the I / O monitoring unit 1231 will be described with reference to FIG. The existing reserved area capacity of the logical volume in each LUN is the size 2400 of each segment in the LUN from the segment management table 2000 of FIG. 2 in which the segment number of the physical logical management table 3000 of FIG. 3 and the size 2400 of each segment are described. Are calculated (step 6001). When an I / O request is sent from the host 1100, the LBA accessed on the logical volume is detected by the I / O request (step 6002). When the area secured is larger than the accessed LBA (step 6003), the I / O processing is requested to the physical logical address management unit 1235 (step 6005). When the area secured from the accessed LBA is smaller (step 6003), a request for securing the capacity is made to the capacity increasing / decreasing unit 1232 (step 6004), and an I / O process is requested to the I / O processing unit 1234 (step 6005). )

  Next, the processing operation of the capacity increasing / decreasing unit 1232 will be described with reference to FIG. The operation differs between a request from the I / O monitoring unit 1231 and a request from the command unit 1132 of the host 1100. First, the operation from the I / O monitoring unit 1231 will be described. When a command to increase capacity is requested from the I / O monitoring unit 1231 (step 7001), volume increase processing (step 7002) is performed, and the processing ends. On the other hand, when a command to reduce the capacity for mLBA is requested from the command unit 1132 of the host 1100 (step 7001), volume reduction processing for mLBA is performed (step 7003).

  In the volume increase processing (step 7002), the following processing is performed in cooperation with the segment management unit 1233 and the physical logical address management unit 1235 (FIG. 8). First, an unused segment acquisition request is sent to the segment management unit 1233 as the segment number parameter SN = −1 (step 8001). In the segment management unit 1233, when the requested segment exists, the segment number parameter SN is returned with the segment number (SN> = 0), and when the requested segment does not exist, SN = −1 is returned (step 8002). . The capacity increasing / decreasing unit 1232 that has received the segment number determines whether or not the segment number parameter SN is a positive number (step 8003). If SN> = 0, the process is continued. If SN <0, an error is returned (step 8006), and the process is terminated. The capacity increase / decrease unit 1232 makes a volume connection request to the physical logical address management unit 1235 based on the segment number SN acquired from the segment management unit 1233 (step 8004). Based on the segment number, the physical logical address management unit 1235 combines the segment with the last segment of the logical volume that is the target LUN to the physical logical management table 3000 (step 8005), and ends the process. Specific operation of the physical logical address management unit 1235 will be described later.

  On the other hand, the volume reduction process (step 7003) is performed as follows (FIG. 9). When the reduced capacity is mLBA, one or a plurality of segment numbers in which no reduced logical volume area exists are acquired from the physical logical management table 3000 (step 9001). As a segment acquisition procedure, first, a reduced LBA is calculated from the physical logical management table 3000 as LBA (END) -m from the end LBA (3400). Based on the reduced LBA, one or more segment numbers (SN) that do not overlap with the reduced LBA are acquired on the target LUN.

  Next, a return request for these segments is made to the physical logical address management unit 1235 (step 9002). The physical logical address management unit 1235 sets the columns (SN 3200, LBA (START) 3300, LBA (END) 3400) matching the segment number from the list of logical volumes specified by the LUN on the physical logical management table 3000 for the received segment. Disconnect (step 9003). The capacity increasing / decreasing unit 1232 instructs the segment management unit 1233 to manage the segment number SN thus separated as an unused segment (step 9004). The segment management unit 1233 manages the separated segment as an unused segment (step 9005). The host 1100 is notified that the logical volume has been disconnected (step 9006). This is the end of the volume reduction processing procedure.

  When the segment management unit 1233 receives an instruction from the capacity increase / decrease unit 1232, the segment management unit 1233 manages the use status of segments described in the segment management table 2000 and the addition / deletion of segments. The management of the segment usage status follows the management procedure of the segment management table 2000 described above.

The operation of managing the segment usage status will be described with reference to FIG. First, a volume increase / decrease request is passed from the capacity increase / decrease unit 1232 together with an argument segment number. When segment number <0, it is determined as an unused segment acquisition request (step 10001), and the processing continues to step 10002. The unused segment acquisition request checks whether or not an unused volume exists using the segment management table 2000 (step 10002). If there is no unused volume (step 10002), it notifies the capacity increasing / decreasing unit 1232 as an error (step 10005). If there is an unused segment (step 10002), the segment number of the unused segment is returned to the capacity increasing / decreasing unit 1232 (step 10003). Then, the usage status corresponding to the segment in the segment management table is changed to 1 indicating in use (step 10004), and the process proceeds to step 10007.
On the other hand, when the segment number> = 0 (step 10001), it is determined that the operation moves to the unused state of the segment, and the processing is continued in step 10006. In step 10006, the usage status corresponding to the segment number is set to 0 which is not used from the segment management table 2000, and the process proceeds to step 10006. In step 10007, the usage rate 290 of the segment management table 2000 is used to calculate the segment usage rate (number of used segments / total number of segments) from the number of used segments and the total number of segments. % Or not is determined. If it is 90% or more, the maintenance staff is requested to add the magnetic disk storage device 1300 (step 10008), and the process is terminated. If the usage rate is 90% or less, the process is terminated as it is. It is assumed that 90%, which is the threshold value of the utilization rate, can be changed variably by the maintenance staff depending on the reliability of the system. This is the end of the description of the volume usage status management operation of the segment management unit 1233.

  Next, operations for adding / deleting the magnetic disk storage device 1300 will be described with reference to FIG. The operation of volume addition / deletion is determined in step 11001. If it is added, the process proceeds to step 11005. If it is deleted, the process proceeds to step 11002. When the magnetic disk storage device 1300 is added, the capacity of the magnetic disk storage device 1300 is confirmed in step 11005 (step 11005). The number and size of segments are determined based on the confirmed capacity (step 11006). The number and size of segments may be fixed values or values specified by maintenance personnel. Based on this value, every time a magnetic disk storage device 1300 is added to the end of the segment management table 2000, the disk ID 2100, the segment number 2200, the start position 2300, the segment size 2400, and the usage status 2500 are not yet used. Inserting 0 representing use (step 11007), the process is terminated.

  On the other hand, in the case of volume deletion, it is determined whether or not the segment with the corresponding segment number is an unused segment (step 11002). If it is an unused segment, the segment number column is deleted from the segment management table 2000 (step 11003). If the segment is not an unused segment (step 11002), it is returned as an error, and the process is terminated (step 11004).

  This is the end of the description of the segment addition / deletion operation of the segment management unit 1233. The segment management unit 1233 performs segment usage status management and addition / deletion operations according to the procedure described above.

  The operation process of the I / O processing unit 1234 will be described with reference to FIG. First, an I / O request sent from the I / O monitoring unit 1231 to the I / O processing unit 1234 requests the physical logical address management unit 1235 to convert the LBA of the I / O request into a physical address (step 12001). . Next, the type of I / O request is determined (step 12002). In the case of read I / O, data is read from the magnetic disk storage device 1300 using the physical address (step 12003), and the read data is returned to the host 1100 (step 12004). In the case of write I / O, data is written to the magnetic disk storage device 1300 using the physical address (step 12005), and the completion of the writing is notified to the host 1100 (step 12006). This is the end of the description of the procedure of the operation of the I / O processing unit 1234.

  The operation of the physical logical address management unit 1235 is divided into two operations. First, a physical logical address conversion procedure requested from the I / O processing unit 1234 will be described. This conversion procedure will be described with reference to FIG. 13, and the structure between the physical and actual physical addresses will be described with reference to FIG. Use the following rules for conversion: First, a segment group for the LUN that is the target of the I / O request is selected (step 13001). The physical disk number and LBA are calculated from the LBA (step 13002).

  First, the LBA target (14300) refers to the start LBA of each segment from the physical logical management table 3000 based on the LBA target described in the I / O request from a plurality of segment groups to which the target LUN belongs. Select the segment number (SN) to access. Using the segment management table 2000 from the segment number SN that has been found, the LBA target (14300) calculates the disk ID to be accessed and the physical address from the beginning of the segment magnetic disk device. Since the address on the segment on which data is operated from the head physical address LBAN (14003) of the Nth segment is LBAtarget-LBAN, the physical address of the segment from the head of the magnetic disk storage device 1300 is (the head LBA LBA segstart of the segment). ) + (LBA-LBAN) (14005). The LBA (start LBA LBA segstart of segment) + (LBA-LBAN) (14005) which is the LBA of the disk ID 2100 and the magnetic disk storage device 1300 with the disk ID 2100 is returned to the I / O processing unit 1234 (step 13003).

  Another operation of the physical logical address management unit 1235 is addition and deletion of the volume to the physical logical management table 3000. This operation will be described with reference to FIG. A LUN in the physical logical management table is selected (step 15001). When a volume is added (step 15002), a segment number added by the capacity increasing / decreasing unit 1232 is given, and a segment is added to the end of the segment column in the LUN volume to be added (step 15003). At this time, as to how the segment added from the host 1100 appears, the LBA has an additional volume between addresses LBAN (14003) and LBAN + SIZE (14004). When deleting a volume (step 15002), after receiving an instruction to delete mLBA from the capacity increase / decrease unit 1232, the capacity of mLBA for each segment unit from the end of the segment column in the LUN volume to be deleted Is deleted (step 15004). The deleted segment returns the segment number to the capacity increasing / decreasing unit 1232 (step 15005). Above, description of operation | movement of the physical logical address management part 1235 is complete | finished.

  The command unit 1132 of the host 1100 reduces the capacity of data in the volume used on the application 1110 side, and then receives an instruction from the application or the user, and sets the storage area of the logical volume provided from the volume server 1200 for each segment unit. Is a program for instructing the volume server 1200 via the I / F 1140. This operation will be described with reference to FIG.

  First, the application 1110 receives a request to reduce the volume of mLBA (step 16001). The command unit 1132 reduces the final block position of the logical partition information of the logical volume managed under the OS 1130 by mLBA (step 16002). The command unit 1132 issues a volume having a capacity of mLBA to the volume server capacity increasing / decreasing unit 1232 (step 16003). The capacity increasing / decreasing unit 1232 performs volume reduction processing (step 16004), and returns the result to the host 1100 side. This is the end of the description of the operation of the command unit 1132.

  Next, a cooperative operation with the application 1110 is shown. As described above, the host 1100 can handle the logical volume provided from the volume server 1200 as an existing volume. However, the volume server 1200 can access a volume that is larger than the segment held by the volume indicated by the LUN. When it occurs, the volume is added from the segment management unit 1233. This operation is the same as the operation of the volume providing unit 1230 described above.

On the other hand, when the capacity of the volume expanded by the application 1110 itself is reduced and the capacity used in the actual application 1110 decreases, the capacity of the difference before and after the reduction is sent to the command unit 1132 of the host 1100 on the application 1110 side. And the above-described volume reduction procedure (FIG. 16) in the volume server via the command unit 1132 of the host 1100 is executed. With these procedures, it is possible to provide an appropriate logical volume capacity in accordance with the operation of the application 1110 with the number of segments being used in the volume server.
<Second Example>
In the first embodiment, when a write I / O access occurs outside the storage area of the logical volume provided from the volume server 1300 by the host 1100, the volume server 1200 allocates the storage area sequentially, thereby The storage area was expanded to the provided logical volume. In the second embodiment, the storage capacity of the logical volume is expanded in advance via the command unit 1300 of the host 1100 according to the instruction of the application 1110, and the OS 1130 can recognize the size of the storage capacity of the logical volume and use it for the application 1110. To. This procedure will be described with reference to FIG.

  First, the command unit 1132 of the host 1000 receives an lLBA storage area expansion request from the application 1110 (step 18001). The command unit 1132 of the host 1100 issues a write I / O to the LBA obtained by adding up the total capacity of the corresponding logical volume and the extended area mLBA. The I / O monitoring unit 1231 of the volume server 1200 performs capacity expansion processing to the capacity increasing / decreasing unit 1232 because the volume is expanded by mLBA from the volume provided by the volume server (step 18003). If the write I / O succeeds and the storage area increases, step 18005 is performed. If the storage area fails, the storage area is not expanded and the process ends. In step 18005, the command unit 1132 of the host 1100 increases the number of end blocks of the logical partition information of the OS 1130 for the corresponding logical volume provided from the volume server 1200 by mLBA. This is the end of the description of the procedure for expanding the storage area of the logical volume provided from the volume server 1200 via the application 1110.

The block diagram of a memory | storage device system. The figure which shows the segment management table 2000. FIG. The figure which shows the physical logical management table 3000. FIG. The figure which shows the logical structure relationship of LUN0 (4000), segment 1 (3410), segment 2 (3420), and segment 3 (3410). An operation flow of I / O processed between the host 1100 and the volume server 1200. The operation flow of the I / O monitoring unit 1231. The operation flow of the capacity increasing / decreasing unit 1232. An operation flow of volume expansion processing among the segment management unit 1233, the capacity increase / decrease unit 1232, and the physical logical address management unit 1235. An operation flow of volume reduction processing among the segment management unit 1233, the capacity increase / decrease unit 1232, and the physical logical address management unit 1235. The operation flow of the segment use condition management of the segment management part 1233. FIG. An operation flow at the time of disk addition / deletion by the segment management unit 1233. The operation flow of the I / O processing unit 1234. The operation | movement flow of physical logical address conversion of the physical logical address management part 1235. FIG. The figure showing the logical structural relationship among LUN 14000, segment 14201 and segment 14202, and magnetic disk storage device 14100. The operation | movement flow at the time of the addition and deletion of the segment to the physical logical management table of the physical logical address management part 1235. An operation flow at the time of expanding the logical volume storage capacity between the command unit 1132 in the host 1100 and the capacity increase / decrease unit 1232 in the volume server 1200. The figure which shows the physical format of an I / O command. An operation flow at the time of expanding the logical volume storage capacity between the command unit 1132 in the host 1100 and the capacity increase / decrease unit 1232 in the volume server 1200.

Explanation of symbols

1100: Host, 1110: Application, 1130: Operating system (OS), 1131: Volume device part, 1132: Command part, 1140: Channel interface, 1200: Volume server, 1210: Channel interface, 1220: Channel interface, 1230: Volume Providing unit, 1231: I / O monitoring unit, 1232: capacity increasing / decreasing unit, 1233: segment managing unit, 1234: I / O processing unit, 1235: physical logical address managing unit, 1236: volume providing unit, 1300: magnetic disk storage apparatus

Claims (5)

An automatic capacity expansion method in a volume providing device connected to at least one host computer and at least one disk storage device is:
Accepts an I / O request from the host computer to the logical volume,
Read the logical block address to be accessed of the I / O request,
When the storage area of the logical block address accessed by the I / O request does not exist on the logical volume, a storage area is allocated from an unused disk storage device and the storage area of the logical volume is dynamically expanded. A method for automatically expanding the capacity of a storage device. An automatic capacity expansion method in a volume providing device connected to at least one host computer and at least one disk storage device is:
Accept a logical volume reduction request from the host computer,
Read the logical block address subject to the reduction request,
An automatic capacity expansion method for a storage device, wherein the storage area of a logical block address designated by the reduction request is reduced.   2. The method for automatically expanding the capacity of a storage device according to claim 1, wherein the storage area of the logical volume is expanded based on an instruction from an application.   3. The method for automatically expanding the capacity of a storage device according to claim 2, wherein the storage area of the logical volume is reduced based on an instruction from an application. The volume providing device connected to at least one host computer and at least one disk storage device is:
Means for receiving an I / O request from the host computer to the logical volume;
Means for reading a logical block address to be accessed by the I / O request;
Means for allocating a storage area from an unused disk storage device and dynamically expanding the storage area of the logical volume when the storage area of the logical block address accessed by the I / O request does not exist on the logical volume; A volume providing apparatus.

Images