WO2016199230A1 - Système de mémorisation - Google Patents

Système de mémorisation Download PDF

Info

Publication number
WO2016199230A1
WO2016199230A1 PCT/JP2015/066647 JP2015066647W WO2016199230A1 WO 2016199230 A1 WO2016199230 A1 WO 2016199230A1 JP 2015066647 W JP2015066647 W JP 2015066647W WO 2016199230 A1 WO2016199230 A1 WO 2016199230A1
Authority
WO
WIPO (PCT)
Prior art keywords
cartridge
file
storage
path
storage device
Prior art date
Application number
PCT/JP2015/066647
Other languages
English (en)
Japanese (ja)
Inventor
昭信 渡邊
千代 大野
小林　正幸
Original Assignee
株式会社日立製作所
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 株式会社日立製作所 filed Critical 株式会社日立製作所
Priority to PCT/JP2015/066647 priority Critical patent/WO2016199230A1/fr
Priority to JP2017522797A priority patent/JPWO2016199230A1/ja
Publication of WO2016199230A1 publication Critical patent/WO2016199230A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F12/00—Accessing, addressing or allocating within memory systems or architectures

Definitions

  • the present invention relates to a storage system using a removable storage device.
  • an optical disk which is one of the removable media, has medium characteristics that can be recorded and stored for a long period of time (for example, several decades) compared to other electronic recording media, and is suitable for use in archives.
  • a device equipped with a mechanism for storing a plurality of optical disk cartridges mounted with a plurality of optical disk drives for reading / writing data from / to such an optical disk is called an optical disk library apparatus.
  • Patent Document 1 information on a source logical unit including data location information on the source logical unit of the old storage system is acquired from a specified source logical unit on the old storage system, and the information on the source logical unit is obtained.
  • the obtained old storage system data is written into the new storage system, and the destination location information is stored in the new storage system.
  • data in the old storage system is migrated.
  • the optical disc library apparatus records on which optical disc the archive file is stored as management information (hereinafter referred to as file management information or metadata).
  • file management information hereinafter referred to as file management information or metadata.
  • the lifetime of the optical disk is longer than the lifetime of the devices constituting the optical disk library apparatus (for example, several years). If the file management information is lost due to the destruction of the optical disk library apparatus, the archive file in the optical disk cannot be read. End up. In addition, with the increase in the capacity of data to be stored and the lengthening of the storage period, not only the data capacity but also the capacity of metadata for managing data increases. If the metadata is stored in the HDD storage device, the cost of the HDD and the migration cost during long-term storage increase. On the other hand, if all metadata is stored on an optical disk, there is a problem that access performance deteriorates.
  • the management information of the file stored in the optical disc designated by the optical disc library apparatus is only the management information of the file stored in the storage medium loaded in the optical disc library apparatus.
  • the recording position of the file is stored in a storage device outside the optical disc library apparatus, and the optical disc in which the file is stored is changed to an optical disc library in response to an access request to the file on the optical disc.
  • the management information has a problem of increasing the consumption of the storage area of the storage device and increasing the cost of the storage device.
  • management information indicating which file is stored on which optical disk in the cartridge is appropriately managed in response to a file read request from the user, thereby reducing the capacity cost of the optical disk and the migration cost of the storage system.
  • Block diagram showing the physical configuration of the optical disc library apparatus Block diagram showing the physical configuration of the buffering server Functional block diagram of a program developed in a local memory in the CPU of the optical disc library apparatus Functional block diagram of the program developed in the local memory in the CPU of the buffering server Management table stored in the control memory in the buffering server Common information for each cartridge Path table for each cartridge Recording position table for each file Account information table Share information table Cartridge information table Processing flow when loading cartridge into optical disk library device Write processing flow Read processing flow Restore processing flow Removal process flow Delete table Pointer table Transfer processing flow from storage device to optical disk Transfer processing flow from optical disk to storage device Hash table for each cartridge using hash Same hash path table Collision table when hashes for file identification collide Hash registration process flow to path table Path table search processing flow using hash Same hash path table Time stamping process flow for file identification Time stamp identification process flow for file identification Recording process flow in the recording time range for each cartridge Process flow for selecting the recording time range for each cartridge Owner ID information Volume information table Disk information table Processing flow when loading cartridge into optical
  • management table various types of information may be described using an expression such as “management table”, but the various types of information may be expressed using a data structure other than a table. Further, the “management table” can be referred to as “management information” to indicate that it does not depend on the data structure.
  • the process may be described with “program” as the subject, but the program is executed by a processor, for example, a CPU (Central Processor Unit), so that a predetermined process is appropriately stored. Since the processing is performed using resources (for example, memory) and communication interface devices (for example, communication ports), the subject of processing may be a processor.
  • the processor may have dedicated hardware in addition to the CPU.
  • the computer program may be installed on each computer from a program source.
  • the program source may be, for example, a program distribution server or a storage medium.
  • each element for example, a logical volume can be identified by a number or the like, but other types of identification information such as a name may be used as long as the information can be identified.
  • an area may be secured in the storage device 104 to supplement the memory capacity, and it may be used in the same manner as the memory.
  • a file system of a buffering server may be used, a physical address range of the storage apparatus may be secured, or the area of the storage apparatus is divided into partitions. Some may be used.
  • FIG. 1 shows a physical configuration of the optical disc library apparatus.
  • an optical disc library apparatus is one apparatus constituting a computer system including, for example, an optical disc library apparatus 101, a buffering server 102, a management server 103, a storage apparatus 104, and a host computer 105.
  • the optical disc library apparatus 101 and the buffering server 102 are connected, for example, via a communication line, and the optical disc library apparatus 101 and the management server 103 are connected, for example, via a communication line.
  • the buffering server 102 and the storage device 104 are connected via a communication line, for example, and the buffering server 102 and the management server 103 are connected via a communication line, for example.
  • the management server 103 and the storage device 104 are connected via a communication line, for example.
  • the buffering server 102 is connected to the network via a communication line.
  • the management server 103 is connected to a network via a communication line.
  • Each host computer 105 is connected to a network via a communication line.
  • the above-described communication line is configured as a wire such as a metal cable or an optical fiber cable.
  • the communication line between the management server 103 and the network, and the communication line between the host 105 and the network may be a common network.
  • the network is a communication network, for example, a SAN (Storage Area Network) or a LAN (Local Area Network).
  • the optical disc library apparatus 101 includes, for example, one or a plurality of CPUs 111, one or a plurality of control memories 112, one or a plurality of transport apparatuses 113, and one or a plurality of storage / reproduction apparatuses (hereinafter referred to as drives). 114 and one or more cartridges 115.
  • Each cartridge 115 contains one or a plurality of optical disks 116, and each cartridge 115 can be detached from the optical disk library apparatus 101.
  • Each optical disc includes at least a Blu-ray Disc (TM) (hereinafter referred to as BD) or a Hologram Disc as a storage device for storing data.
  • TM Blu-ray Disc
  • the optical disc may be another type of physical storage device instead of or in addition to at least one of them.
  • the one or more optical disk transport devices 113 and the one or more drives 114 are connected to the optical disk library device via a communication path such as a copper cable, an optical cable, or a bus.
  • one or more RAID (Redundant Array of Independent Disks) groups can be configured with a plurality of optical disks.
  • the optical disc library apparatus 101 performs disc transport processing for the optical disc transport device 113 or data input / output processing for the drive 114, that is, data write (write) or read (read) to the optical disc. ) To control.
  • the command transmitted from the buffering server 102 to the optical disc library 101 apparatus is from the host computer that is the access request source of data write or read to the original optical disc. It will be described later.
  • the optical disk library apparatus 101 provides the optical disk loaded in the drive 114 to the buffering server 102 that is the command transmission source, with the real storage area already allocated and as a logical volume to be accessed. To do.
  • the buffering server 102 can refer to or identify the actual storage area of the optical disk loaded in the drive 114 by, for example, Logical Block Address (hereinafter, LBA).
  • LBA Logical Block Address
  • the optical disc library apparatus 101 includes, for example, a storage resource, a communication interface device (hereinafter, the interface device is abbreviated as “I / F”), and an arithmetic device connected thereto.
  • a storage resource for example, a storage resource, a communication interface device (hereinafter, the interface device is abbreviated as “I / F”), and an arithmetic device connected thereto.
  • I / F communication interface device
  • the arithmetic unit is, for example, a CPU (Central Processor Unit).
  • the CPU can store a microprogram described later in a volatile local memory inside the CPU.
  • the main body of the microprogram is stored in a non-volatile memory area in the CPU in case the power supply to the CPU is lost.
  • the storage resource is, for example, a control memory.
  • the control memory is, for example, a DRAM (Dynamic Random Access Memory) that loses stored data (volatile) if power is not supplied.
  • the control memory is, for example, a flash package containing a (nonvolatile) flash memory chip that can hold data without power supply.
  • the control memory may be configured by combining a volatile memory and a nonvolatile memory.
  • the communication I / F includes, for example, a buffering server I / F that receives reads and writes from the buffering server, and a management server I / F that receives management commands from the management server.
  • These buffering server I / F, management server I / F, CPU, control memory, optical disk transport device 113, and drive 114 are connected to each other via a communication line such as a bus.
  • the hardware configuration of each host 105 and management server is the same as that of a general computer. That is, each host 105 and management server 103 includes a communication interface device, a storage resource, and an arithmetic device connected to them.
  • An example of the communication interface device is a host bus adapter (HBA) for performing communication via a network.
  • the storage resource is composed of, for example, a memory or a built-in HDD.
  • FIG. 2 shows the physical configuration of the buffering server.
  • the buffering server is one device constituting a computer system including, for example, the buffering server 102, the optical disc library apparatus 101, the management server 103, the storage apparatus 104, and the host computer 105. Further, the connection between the buffing server 102 and other computers and devices is as described above.
  • the buffering server (102) includes, for example, one or more CPUs and one or more control memories.
  • the buffering server 102 transmits a data input / output command to the library device or a disk transport processing command to the optical disk transport device 113 according to the file storage / read instruction received from the host computer 105, and receives a response. That is, data is written (written) or read (read) from the optical disk via the optical disk library apparatus.
  • the buffering server 102 provides the optical disk loaded in the drive 114 from the optical disk library apparatus as a logical device or logical volume to which a real storage area has already been allocated.
  • the real storage area is referred to by an LBA. Or can be identified.
  • the host computer 105 is provided with the above-described logical device or logical volume as a file system from the buffering server 102, and can refer to or identify the real storage area as a file path, for example. it can.
  • the unit of input / output performed by the host computer 105 for the buffering server 102 is a file.
  • the host computer 105 only reads a meaningful data collection as a file, and if the host computer can identify this data collection, the file is, for example, a block, record, or file path. For example, an ID number may be used.
  • the host computer 105 is operated by a user, and a user performs a file storage command or a file read command for the buffering server 102 described later through the host computer 105.
  • the buffering server has, for example, a storage resource, a communication interface device, and an arithmetic device connected to them.
  • the arithmetic unit is, for example, the CPU 211.
  • the storage resource is, for example, the control memory 212.
  • the control memory is, for example, a DRAM (Dynamic Random Access Memory) that loses stored data (volatile) if power is not supplied.
  • the control memory is, for example, a flash package containing a (nonvolatile) flash memory chip that can hold data without power supply.
  • the control memory may be configured by combining a volatile memory and a nonvolatile memory.
  • the control memory may have a configuration in which a volatile memory and a part of the storage area of the storage device 104 are combined, and the capacity of the memory having a high bit cost can be reduced.
  • a part of the storage area of the storage device 104 may be secured as a virtual drive area having the same capacity as the capacity of the optical disk.
  • the optical disk By recording the recording position table 513 and the pointer table 521 in the virtual drive, the optical disk can be recorded.
  • the writing process is a bit-by-bit sequential writing from the virtual drive to the optical disk, so that the process is simple and there is an effect that no seek occurs and the transfer is fast.
  • by securing a plurality of virtual drives for example, when users of a plurality of accounts access simultaneously, if a virtual drive is allocated for each account, it becomes easy to control the optical disk writing of metadata in units of accounts. There is an effect.
  • the communication I / F for example, a host I / F that receives a file storage / read instruction from the host computer, a library I / F that transmits read / write to the optical disc library apparatus, and a management command from the management server A management server I / F, and a storage I / F that reads from and writes to the storage device.
  • the host computer (105) includes an operating system (OS) and an application (APP) executed on the OS.
  • OS operating system
  • APP application
  • OS operating system
  • APP application
  • the management server (103) is composed of an operating system (OS) and management software executed on the OS.
  • the management software uses the input / output interface provided by the OS to change the configuration of the buffering server 102, the optical disc library apparatus 101, and the storage apparatus 104 according to instructions from the system administrator, and to manage the status in the apparatus by system management Displayed to the user.
  • a plurality of management software may be operating on different OSs.
  • a plurality of management software may be operating on different hosts.
  • the buffering server 102 When the buffering server 102 receives a file storage command from any one of the host computers 105 via the host I / F, the buffering server 102 stores the file received from the host computer 105 in the storage device 104.
  • the buffering server 102 reads the file stored in the storage device 104 via the storage I / F, and writes the file to the optical disc via the drive 114 in the optical disc library device 101 via the library I / F.
  • the procedure in which the buffering server 102 writes the file received from the host to the optical disc via the drive 114 in the optical disc library apparatus 101 is as follows.
  • the buffering server 102 checks the volume number (A) and the volume address (B) in which there is a space larger than the file size received from the host, based on the volume information table 517 described later. If there is no available space, the system administrator may be notified via the management server 103 to insert an unused cartridge.
  • the buffering server 102 takes out the optical disk corresponding to the corresponding disk number from the cartridge to the optical disk transport device 113 based on the volume number (A) and the disk information table 524 described later.
  • the drive 114 is instructed to be conveyed.
  • a write command is transmitted to the optical disc library apparatus 101 using the start address (B) of the recordable area of the optical disc, the file size (C), and the file received from the host computer 105 as parameters.
  • the optical disc library apparatus 101 Upon receiving a write command from the buffering server 102, the optical disc library apparatus 101 writes a file to the optical disc stored in the drive 114 based on the address (B) and file size (C) included as parameters in the write command. The write completion is notified to the buffering server 102.
  • the buffering server 102 When receiving the write completion notification, the buffering server 102 updates the used capacity in the volume information table 517 based on the written file size.
  • the buffering server 102 may delete the file that has been stored on the optical disc from the storage device 104.
  • the buffering server 102 stores the file path and file size included in the file storage command received from the host 105, the logical volume number and logical address in which the file is stored. These pieces of information existing for each file stored on the optical disc are collectively referred to as stub information. Further, a cartridge information table for identifying a cartridge included in the optical disk storing the file, an account information table for identifying an account that owns the file and a share that is an allocated storage area, and a share information table are stored.
  • the stub information includes common information for each cartridge described later, a path table for each cartridge, a recording position table for each file, a hash table, and a collision table.
  • the buffering server 102 may be configured to notify the host computer 105 that processing of the file storage command has been completed when the file is stored in the storage device 104, or the optical disk library device 101.
  • the configuration may be such that it is performed when a file write completion notification is received from and the stub information is created.
  • the former is generally used because the file read from the storage device 104 is stored in the drive 114 in the optical disk library device 101 rather than the process of storing the file received from the host computer 105 in the storage device 104. This is because the process of writing to the optical disc via the optical disc is much slower.
  • the above is the basic operation of the buffering server 102 and the optical disc library apparatus 101 when a file storage command is received from the host.
  • the buffering server 102 stores the volume number (A), the recording address (B), the file included in the stub information that can be uniquely identified by the file path. Get the size (C). If the file path is not stored as part of the stub information, the failure of the file read command received from the host computer 105 may be notified to the system administrator via the management server 103.
  • the buffering server 102 stores the file to be read based on the volume number (A) and the disk information table 524 described later. Get the cartridge number. Based on the online / offline attribute information in the disk information table 524, the buffering server 102 obtains information on whether or not a cartridge corresponding to the cartridge number exists in the optical disk library apparatus. When a cartridge corresponding to the cartridge number exists in the optical disc library apparatus 101, the disc number storing the file to be read is acquired based on the volume number (A) and the disc information table 524. The optical disk transport device 113 is instructed to take out the optical disk corresponding to the disk number from the cartridge and transport it to the drive 114. The buffering server 102 transmits a read command to the optical disc library apparatus 101 using the recording address (B) and the file size (C) as parameters to the optical disc library apparatus.
  • the optical disc library apparatus 101 reads a file from the optical disc stored in the drive 114 based on the volume address (B) and the file size (C) included as parameters in the read command, and transmits the file to the buffering server 102. Further, the optical disc library apparatus 101 notifies the buffering server 102 of the completion of reading when the transmission of the file to the buffering server 102 is completed.
  • the buffering server 102 When the buffering server 102 receives the read completion notification, the buffering server 102 transmits the file read from the optical disc library apparatus to the host computer 105, and notifies the completion of the file read command when the transmission is completed.
  • the above is the basic operation of the buffering server 102 and the optical disc library apparatus 101 when a file read command is received from the host.
  • FIG. 3 shows a functional block diagram of a program developed in the local memory in the CPU of the optical disc library apparatus 101.
  • the program of the optical disc library apparatus 101 includes a server command control unit 311 and a resource management unit 312.
  • the server command control unit 311 When the server command control unit 311 receives a write command or read command from the buffering server 102 via the buffering server I / F, the logical address in the logical volume specified by the parameter of the write command or read command To write or read data for the specified size.
  • the resource management unit 312 monitors the cartridge loading status in the optical disc library apparatus, notifies the management server of the loaded cartridge or the disc inside the cartridge (disc type, capacity, free capacity, attributes), and If it is not registered in the information table 524, it is registered.
  • FIG. 4 shows a functional block diagram of a program developed in the local memory in the CPU of the buffering server 102.
  • the program of the buffering server 102 includes a host command control unit 411, a recording position management unit 412, and a migration control unit 413.
  • the host command control unit 411 When the host command control unit 411 receives a file storage command from the host computer 105 via the host I / F, the host command control unit 411 stores the file in the file path specified by the parameter on the storage apparatus 104.
  • the host command control unit 411 When the host command control unit 411 receives a file read command from the host computer 105 via the host I / F, the host command control unit 411 is based on the file stub information existing in the control memory of the storage apparatus 104 or the buffering server 102. The volume number, volume address, and file size are acquired, and a read command is issued to the optical disc library apparatus using these as parameters. When a file is read from the optical disk library device, it has a function of transmitting it to the host computer 105.
  • the recording position management unit 412 performs processing such as recording, updating, and reading management information stored in the control memory in the buffering server.
  • the management information is written to and read from the storage device 104.
  • the migration control unit 413 has a function of writing a file received from the host computer 105 and stored in the storage apparatus 104 to the optical disc library apparatus 101.
  • FIG. 5 shows a management table stored in the control memory in the buffering server (102).
  • FIG. 6 shows the configuration of common information 511 for each cartridge.
  • the version information 611 indicates the version of the standard that defines the format of management information, the hash calculation method, and the like.
  • the oldest recording time 612 indicates the oldest time among the recording times of the files recorded on the cartridge.
  • the latest recording time 613 indicates the newest time among the recording times of the files recorded on the cartridge.
  • the number of files 614 indicates the total number of files recorded on the cartridge.
  • the size 615 indicates the cumulative size of files recorded on the cartridge.
  • FIG. 7 shows the configuration of the path table 512 for each cartridge.
  • the path number 711 indicates the total number of paths of files and directories included in the cartridge.
  • the reserve 712 stores a fixed value and does not function in the current version, but is a reserved field for use when expanding functions in a future version.
  • the storage location 713 of the file #N recording position table is information relating to the path of the Nth file or directory included in the path table 512, and includes a path, a directory / file identification flag, a recording position table pointer, and a reserve which will be described later.
  • the path 717 is a field for storing the path of the file #N. Assuming a maximum path length of 1024 bytes, the field length is 1024 bytes.
  • the directory / file identification flag 716 is a field for storing 0 if the file #N is a file and 1 if the file #N is a directory.
  • the recording position table pointer 715 is a field for storing an offset value from the head in a recording position table for each file described later.
  • the recording position number 718 indicates the number of recording position tables for each file pointed to by the recording position table pointer 715.
  • Reserve 714 stores a fixed value and does not function in the current version, but is a reserved field for use when expanding functions in a future version.
  • FIG. 8 shows the configuration of the recording position table 513 for each file.
  • the directory number 811 is a field for storing the total number of directories recorded in the cartridge.
  • the reserve 812 stores a fixed value and does not function in the current version, but is a reserved field for use when expanding functions in a future version.
  • the directory # 1 recording position 814 indicates information related to one of the directories recorded on the cartridge, and includes a directory name, a recording time, a size, and an attribute.
  • the directory name 815 is a field for storing a directory name as a character string.
  • the recording time 816 is a field for storing the time when the directory # 1 is recorded.
  • Size 817 is a field for storing the size of directory # 1.
  • Attribute 818 is a field for storing the attribute of directory # 1. For example, attributes such as read-only / rewritable are stored in the bit field.
  • the file number 821 is a field for storing the total number of files recorded on the cartridge.
  • the reserve 822 stores a fixed value and does not function in the current version, but is a reserved field for use when expanding functions in a future version.
  • the file # 1 recording position 823 indicates information related to one of the files recorded on the cartridge, and includes a file name, a recording time, a size, and an attribute.
  • the file name 824 is a field for storing the file name as a character string.
  • Recording time 825 is a field for storing the time when file # 1 was recorded.
  • Size 826 is a field for storing the size of file # 1.
  • Attribute 827 is a field for storing the attribute of file # 1. For example, attributes such as read-only / rewritable are stored in the bit field.
  • the buffering server 102 records the file in the flow as shown in FIG. 12, and as shown in FIG. Read the file in a simple flow.
  • FIG. 9 shows the configuration of the account information table 514 managed by the buffering server.
  • Account name 911 is a field for storing the account name used by the user as a character string.
  • User ID 912 is a field for storing an ID for managing an account corresponding to the account name 911.
  • the group ID 913 is a field for storing an ID for managing the group to which the account belongs.
  • Account belongs to one of the groups and has a unique user ID within the group.
  • FIG. 10 shows the configuration of the share information table 515 managed by the buffering server.
  • Share is an area obtained by dividing the entire recording area of the optical disc library apparatus in units of cartridges. The area size of each share can be set by the user.
  • the owner ID 1011 is a field for storing a share identification ID assigned to each share.
  • the share name 1012 is a field for storing the share name assigned to the share as a character string.
  • the number of accounts 1013 is a field for storing the number of accounts registered for each share permitted to access the share.
  • the account pointer 1014 is a field for storing a link to a list of account names permitted to access the share.
  • the account pointer 0x20001000 points to the share registration account name 1015
  • the account pointer 0x20001040 points to the share registration account name 1016.
  • the share registration account name 1015 is a list of account names permitted to access the share. In FIG. 10, three accounts of “sales_dept_0”, “sales_dept_1”, and “randd_dept_0” are registered for the share with the share name “Japan”.
  • the share registration account name 1016 is a list of account names permitted to access the share. In FIG. 10, three accounts of “sales_dept_0”, “sales_dept_1”, and “randd_dept_1” are registered for the share with the share name “US”.
  • FIG. 11 shows the configuration of the cartridge information table 516 managed by the buffering server.
  • the cartridge ID 1111 is a field for storing a unique ID assigned to each cartridge accessed by the optical disc library apparatus.
  • the owner ID 1112 is a field for storing an owner ID for identifying the share to which the cartridge is assigned. If the owner ID 1112 matches the owner ID 1011 in FIG. 10, it can be regarded as the same owner ID.
  • the online 1113 is a field for storing a flag for identifying whether the cartridge is online or offline. Identification is possible by setting 1 if online and 0 if offline.
  • the total capacity 1114 is a field for storing the total capacity of the recording area of the cartridge.
  • the remaining capacity 1115 is a field for storing the total remaining amount of the recording area of the cartridge, that is, the capacity of the unrecorded area.
  • FIG. 32 shows the configuration of the volume information table 517.
  • the volume information table 517 includes a volume number field (3211), a volume capacity field (3212), and a used capacity field (3213).
  • the volume number is information for uniquely identifying a logical volume as an access target to which a real storage area has already been allocated, and each entry in the volume number field stores a file received from the host computer 105. Stores the volume number of the logical volume.
  • the volume capacity is information indicating the capacity of the logical volume, and each entry in the volume capacity field stores the capacity of the volume corresponding to the logical volume number described above (for example, the unit is Byte).
  • the used capacity is information indicating the capacity of the area where data has been written by the file storage command in the logical volume, and each entry in the volume capacity field indicates the written area of the volume corresponding to the logical volume number described above.
  • the capacity (for example, the unit is Byte) is stored.
  • the volume information table may be stored in the control memory of the buffering server 102, or may be stored in a predetermined optical disc and the control memory or CPU of the buffering server 102 through the optical disc library apparatus 101 as necessary. It may be read into internal local memory.
  • FIG. 18 shows a pointer table.
  • the online cartridge pointer table 520 includes a cartridge ID 1811 and a pointer 1812.
  • the cartridge ID 1811 is a field in which the cartridge ID of an online cartridge loaded in the optical disc library apparatus is set.
  • the pointer 1812 is expanded on the control memory in the buffering server.
  • This is a field for setting a pointer to a memory area in which the cartridge information table 516, the volume information table 517, the deleted share information table 518, and the deleted cartridge information table 519 are stored.
  • the offline cartridge pointer table 521 includes a cartridge ID 1813, a metadata cartridge ID 1814, a disk ID 1815, a recording position size 1816, and a pointer 1817.
  • the cartridge ID 1813 is a field for setting the cartridge ID of a cartridge that has been written or a cartridge that has been registered. For offline cartridges.
  • the metadata cartridge ID 1814 is a field for setting a cartridge ID of a cartridge including an optical disc in which metadata related to the cartridge with the cartridge ID is recorded.
  • the disk ID 1815 is a field for setting the disk ID of the optical disk on which the metadata related to the cartridge with the cartridge ID is recorded.
  • the start position size 1816 is a field for setting the position and size of the stub information file recorded on the optical disc.
  • a pointer 1817 is developed on the control memory in the buffering server.
  • This is a field for setting a pointer to a memory area in which the cartridge information table 516, the volume information table 517, the deleted share information table 518, and the deleted cartridge information table 519 are stored.
  • the pointer table list 525 in which the pointer table is also offline includes a metadata cartridge ID 1818, a disk ID 1819, a recording position size 1820, and a minimum OID maximum OID 1821.
  • the metadata cartridge ID 1818 is a field for storing the cartridge ID of the cartridge including the optical disc on which the pointer table 521 is recorded.
  • Disc ID 1819 is a field for storing the optical disc ID of the optical disc on which the pointer table 521 is recorded.
  • the start position size 1816 is a field for setting the position and size of the pointer table 521 recorded on the optical disc.
  • the minimum OID maximum OID 1821 is a field for storing the maximum value and the minimum value of the share OID included in the pointer table 521.
  • FIG. 17 is a diagram showing a deletion share information table and a deletion cartridge information table.
  • the deleted share information table 518 includes a share name 1711 of the share.
  • the share name 1711 is a field for storing a character string of the share name of the deleted share when the registered share is deleted.
  • the deleted cartridge information table 519 includes a cartridge ID 1712 of the cartridge.
  • the cartridge ID 1712 is a field for storing the cartridge ID of the deleted cartridge when the registered cartridge is deleted.
  • FIG. 33 shows a disk information table.
  • the disk information table 524 includes a volume number 3311, a disk number 3312, and a cartridge ID 3313.
  • the volume number 3311 is information for uniquely identifying a logical volume as an access target to which a real storage area has already been allocated. Each entry in the volume number field stores a file received from the host computer 105. The volume number of the existing logical volume is stored.
  • the disk number 3312 is a unique product number uniquely assigned to each disk when the disk product is shipped.
  • the cartridge ID 3313 is a unique product number uniquely assigned to each cartridge when the cartridge product is shipped.
  • the cartridge has a storage area such as a flash memory, and can hold the cartridge ID, the number of stored disks, and information (capacity, remaining amount, type) of the stored disk.
  • the optical disc library apparatus accesses this storage area and can read and rewrite information.
  • wireless communication may be used, or communication may be performed by connecting via a connector.
  • the hash table 522 and the collision table 523 will be described later.
  • FIG. 12 shows a processing flow when a cartridge is loaded in the optical disc library apparatus.
  • the system administrator loads a cartridge containing one or a plurality of optical discs into an empty slot of the optical disc library apparatus.
  • the resource management unit 312 in the program of the optical disc library apparatus reads information such as the cartridge ID, the total capacity, the total remaining amount, the capacity for each optical disc, and the remaining amount from the cartridge to the management server 103 and performs buffering. This is notified to the recording position management unit 412 of the server.
  • the recording position management unit 412 of the buffering server 102 checks whether it has been loaded in the past in the optical disc library apparatus and registered in the cartridge information table 516 described above.
  • a unique product number assigned to each cartridge when the cartridge product is shipped is registered in the cartridge ID in the cartridge information table 516 described above. Decide on.
  • a method may be used in which a unique product number assigned to each disk at the time of shipping the disk product is registered as a disk number. (S1203)
  • the recording position management unit 412 registers the cartridge number, total capacity, and total remaining amount of the loaded cartridge in the table.
  • the recording position management unit 412 registers the disk number of the unregistered disk in the loaded cartridge in the volume information table 517 described above, acquires a logical volume number that is not used in the table, and assigns it to the disk number. Register in the table. (S1204) Finally, the recording position management unit 412 sets information corresponding to online in the online / offline attribute of the cartridge corresponding to the loaded cartridge number on the cartridge information table 516 based on the loaded cartridge number. (S1204) The management server 103 displays the latest table information after cartridge loading on the management screen and notifies the system administrator. (S1205) FIG. 34 shows an abnormal process flow in the process flow when the cartridge is loaded into the optical disc library apparatus.
  • the resource management unit 312 in the program of the optical disk library apparatus On the other hand, processing for acquiring information such as cartridge ID, total capacity, total remaining amount, capacity for each optical disk, and remaining amount from the cartridge is performed from the cartridge. If the information acquisition processing fails, abnormal processing is performed. When the information acquisition process is successful, the recording position management unit 412 of the buffering server is notified.
  • the recording position management unit 412 of the buffering server 102 checks whether or not the cartridge is registered in the cartridge information table 516 described above. If the cartridge is not registered, the cartridge number, the total capacity, and the total remaining amount of the cartridge are checked. The amount is registered in the table, but if registration fails, abnormal processing is performed. When registration is successful, information corresponding to online is set in the online / offline attribute of the cartridge corresponding to the cartridge number of the cartridge on the cartridge information table 516.
  • the management server 103 displays the latest table information after cartridge loading on the management screen and notifies the system administrator.
  • the management server 103 may display error information on the management screen and notify the system administrator.
  • S3405 By performing abnormality processing, the system administrator can quickly recognize the abnormality, shortening the time required to request subsequent maintenance work, and minimizing the impact on business continuity due to interruptions There is.
  • ⁇ Writing data to optical disc> Data writing to the optical disk and various information table setting processes are triggered when the total capacity of the file written from the host computer 105 stored in the storage device 104 exceeds a certain value (for example, the capacity of the optical disk). To do. Further, even when the total capacity of the file does not exceed a certain value, it may be performed when a certain period (for example, one day) has been exceeded since the file was written. Moreover, it is good also considering the timing which makes a cartridge offline.
  • a certain value for example, the capacity of the optical disk
  • FIG. 13 shows a flow of writing data to the optical disc and setting various information tables.
  • the migration control unit 413 in the program of the buffering server 102 refers to the volume information table 517 and selects a volume having sufficient free capacity.
  • the free capacity is, for example, the capacity obtained by subtracting the used capacity from the volume capacity, and the volume having sufficient free capacity is, for example, the sum of the sizes of one or more files to be written to the optical disk. Means a volume with a free capacity of.
  • the migration control unit 413 may notify the management server 103 to that effect and display it to the system administrator.
  • the migration control unit 413 If there is a volume with sufficient free capacity, the migration control unit 413 reads one or more files stored in the storage apparatus 104 via the storage I / F. (S1302) Next, the migration control unit 413 acquires a volume number (A) and a used capacity (volume address (B)) of a volume having sufficient free capacity based on the volume information table 517.
  • the migration control unit 413 reads the volume number (A), the volume address (B), the sum of the sizes of one or more files read from the storage device 104 (file size (C)), and the read from the storage device 104.
  • a write command is transmitted to the optical disc library apparatus 101 using one or more files as parameters.
  • the server command control unit 311 in the program of the optical disc library apparatus 101 receives a write command from the buffering server 102, the volume number (A) included as a parameter in the write command, a volume information table 517, and a disc information table 524, the optical disk transport device 113 is instructed to take out the optical disk corresponding to the corresponding disk number from the cartridge and transport it to the drive 114.
  • the server command control unit 311 writes the file to the optical disk (disk ID is Z) stored in the drive 114 based on the volume address (B) and the file size (C), and the writing to the buffering server 102 is completed. To be notified.
  • the migration control unit 413 in the program of the buffering server 102 updates the used capacity in the volume information table 517 based on the written file size (C).
  • the used capacity after the update is obtained by adding the file size (C) to the used capacity before the update. That is, the used capacity means a start address where a file can be written next.
  • the migration control unit 413 may delete one or more files written to the optical disc from the storage apparatus 104.
  • the migration control unit 413 applies the following processing to one or more files written to the optical disc.
  • the buffering server 102 includes a volume number (A), a file size (C), a volume address (B) where the file is stored, and a file system provided by the buffering server 102 to the host computer 105.
  • the storage location (file path (D)) of the upper file is stored as a recording position table 513 for each file.
  • the migration control unit 413 searches the path table 512 for each cartridge, compares the storage location path 717 of the recorded file #N recording position table with the file path (D), and confirms that there is no overlapping path. After confirming, a new storage location of the recording position table is added, the file path (D) is stored in the path 717, and 1 or 0 is set to the directory / file identification flag 716 depending on whether it is a directory or a file Set.
  • the migration control unit 413 newly adds a recording position field to the recording position table 513 for each file, and sets a pointer to the added recording position field in the recording position table pointer 715.
  • the migration control unit 413 adds the disc ID 824, the start location 825, the size 826, and the attribute 827 of the added recording location field to the Z of the disc ID, the used capacity before update, the file size (C), and the Read Only. Set attributes.
  • the attribute can be set by assigning a bit field in addition to Read Only.
  • the migration control unit 413 sets the path table 512 for each cartridge and the recording position table 513 for each file recorded in the volume (A) to the volume number (A) as a stub information file for the volume (A). You may store in the corresponding volume (namely, optical disk) by the method similar to the write processing of the file mentioned above. At this time, the process of (S1301) described above is performed so that the sum of the size of one or more files written to the volume (A) and the size of the stub information file does not exceed the optical disk capacity. (S1304) The migration control unit 413 updates the latest recording time 613 of the common information 511 for each cartridge, updates the oldest recording time 612 if necessary, and updates the recorded file by adding it to the number of files 614.
  • FIG. 35 shows an abnormal process flow in the flow of writing data to the optical disc and setting various information tables.
  • the migration control unit 413 refers to the volume information table 517 and tries to select a volume having sufficient free capacity. However, if there is no volume having sufficient free capacity, abnormal processing is performed. .
  • abnormal processing is performed.
  • the migration control unit 413 If there is a volume with sufficient free space, the migration control unit 413 reads one or more files stored in the storage device 104 via the storage I / F, but a file read error occurs. If it does, abnormal processing is performed.
  • the migration control unit 413 transmits a write command to the optical disc library apparatus 101. However, if a file write error occurs, the migration control unit 413 performs a retry process, and performs an abnormal process when the number of retry processes reaches a preset upper limit. I do.
  • the retry process may be to register an optical disc with a write error as a recording / playback volume, discard it, select a new unrecorded volume, and re-record the recorded file in the discarded volume. It is done.
  • a file with a write error is discarded as a recording failure, but a file that has already been recorded before that time can be read normally and the recorded value matches the read value. It can be considered that the file is regarded as a valid file, writing to the volume is terminated, a new unrecorded volume is selected, and the file in which the write error has occurred is re-recorded in the new volume.
  • the management server 103 may display error information and information related to the retry process on the management screen and notify the system administrator.
  • the migration control unit 413 executes (S1304) and (S1305), displays the cartridge information table after writing on the management screen, and notifies the system administrator.
  • the management server 103 may display error information on the management screen and notify the system administrator.
  • S3505 By performing abnormality processing, the system administrator can quickly recognize the abnormality, shortening the time required to request subsequent maintenance work, and minimizing the impact on business continuity due to interruptions There is.
  • the above is the process of writing data to the optical disc and setting various information tables.
  • FIG. 14 shows a flow of processing for reading a file stored on the optical disk of the host computer 105.
  • the file path of the target file to be read is input from the host computer 105, and the share name included in the file path is extracted by analyzing the input file path.
  • the owner ID assigned to the share name can be acquired based on the share name.
  • the cartridge ID assigned to the owner ID can be acquired based on the owner ID. (S1401) With reference to the pointer table 520, it is confirmed whether or not the cartridge ID 1811 is registered. If it is registered, it is recognized that the cartridge is online. If it is not registered, the cartridge is not online or needs to be newly registered. I understand.
  • NULL is set in the pointer 1812, it can be determined that the path table 512 for each cartridge is not online. If a value other than NULL is set, the value is set by the pointer to the path table 512 for each cartridge. It can be determined that it is online. (S1402) If it is not online, the pointer table 521 is referenced to obtain the metadata cartridge and the disk ID and the recording position that match the cartridge ID, the optical disk is loaded into the drive, and the path table and the recording position table are read from the optical disk.
  • (S1403) Search the path table, acquire the storage position 713 where the character string of the path 717 matches, acquire the recording position of the file from the recording position table pointer 715, acquire the disk ID, start position, and size, Specify the optical disc and recording position.
  • (S1404) If it is online, the path table 512 for each cartridge is acquired by following the pointer. Based on the file path, the path table for each cartridge is searched, and if the character string of the path 717 matches, the storage location 713 of the recording position table is referenced.
  • the file recording position 823 is acquired from the recording position table pointer 715, the disk ID 824, the start position 825, and the size 826 are acquired, and the optical disk and the recording position are specified.
  • FIG. 36 shows an abnormal process flow in the process of reading a file stored on the optical disk of the host computer 105.
  • an abnormal process is performed.
  • the cause of the acquisition error may be that the management information to be referenced is lost and does not exist.
  • S3601 When the path table 512 for each cartridge is not online, the optical disk is transported in order to read the path table and the recording position table from the optical disk. However, if the transport fails, abnormal processing is performed.
  • S3603 When the transport is successful, the path table is searched and an optical disk and a recording position are specified, but if the corresponding management information is not found, an abnormal process is performed.
  • the cause of the acquisition error may be that the management information to be referenced is lost and does not exist.
  • FIG. 15 shows a flow of restoration processing of the recording position table 513 from the optical disc.
  • FIG. 15 when the path table for each cartridge is online, no particular processing occurs because the cartridge is already online.
  • S1501 If the path table for each cartridge is not online (S1501), the pointer table 521 is referred to, the metadata cartridge and the disk ID matching the cartridge ID and the recording position are acquired, the optical disk is loaded into the drive, the path table and Read the recording position table from the optical disk.
  • FIG. 37 shows an abnormal process flow in the restore process flow of the recording position table 513 from the optical disk.
  • the optical disk is transported to load the optical disk on which the recording position table is recorded into the drive. Do. (S1502) If no transport error has occurred and the corresponding management information has been restored successfully, the restored management information is displayed on the management screen and notified to the system administrator. (S3705) As an abnormal process, the management server 103 may display error information on the management screen and notify the system administrator. (S3706) By performing abnormality processing, the system administrator can quickly recognize the abnormality, shortening the time required to request subsequent maintenance work, and minimizing the impact on business continuity due to interruptions There is.
  • the restoration processing of the recording position table 513 for each directory from the optical disk may be performed when the computer system replaces a device other than the optical disk, such as the buffering server and the buffering server 102.
  • a device other than the optical disk such as the buffering server and the buffering server 102.
  • FIG. 16 shows a processing flow performed when the cartridge is removed from the optical disc library apparatus.
  • the cartridge to be removed is recorded depending on whether the number of passes 711 is 0 or a positive number. Determine if a disk is included.
  • the path table 512 for each cartridge is identified from the pointer pointed to by the cartridge ID of the cartridge to be removed, written to the optical disk as a stub information file, and the recording position for each file pointed to by the path table 512 for each cartridge
  • the table 513 is written to the optical disc as a stub information file, the pointer table 520 is updated, the pointer of the cartridge ID to be removed is set to NULL, indicating that it is offline, the pointer table 521 is updated, and the cartridge ID 1813 is set.
  • the cartridge ID of the cartridge to be removed is added, and the cartridge ID of the cartridge including the optical disk on which the path table 512 for each cartridge and the recording position table 513 for each file are written
  • the disk ID of the optical disk is recorded in the metadata cartridge ID 1814 and the disk ID 1815, the start position and size written on the optical disk are recorded in the start position size 1816, and the pointer 1817 is stored in the path table 512 for each cartridge. Record the pointer. (S1602) If it is not stored, no update process occurs.
  • FIG. 38 shows an abnormal processing flow in the processing flow performed when the cartridge is removed from the optical disc library apparatus.
  • the cartridge to be removed includes a recorded disk (S3801). If the recorded cartridge includes a recorded disk, the stub information file is written on the optical disk. If a write error occurs, an abnormal process is performed. (S3802) If a recorded disc is not included, an unrecorded cartridge is to be removed, so that no update process occurs, and the cartridge information after removal is displayed and notified to the system administrator. (S3803) As an abnormal process, the management server 103 may display error information on the management screen and notify the system administrator. (S3804) By performing abnormality processing, the system administrator can quickly recognize the abnormality, shortening the time required to request subsequent maintenance work, and minimizing the impact on business continuity due to interruptions There is.
  • the processing flow performed when removing the cartridge from the optical disc library apparatus is to prompt the system administrator via the management server as described in (S1303), for example, and load the corresponding cartridge into the optical disc library apparatus. In order to make a free space in the optical disc library apparatus, it is performed when another cartridge already loaded is removed. Further, as another case, for example, when a system migration accompanying device replacement is necessary, it is performed when a cartridge including an optical disk is removed from the migration source computer system.
  • one or a small number of optical disks 116 in the cartridge 115 is an optical disk dedicated to metadata, the other is an optical disk dedicated to user data, and metadata of user data in the cartridge is stored.
  • a method can be considered in which recording is performed on an optical disk dedicated for metadata in the same cartridge. For example, when 250 optical disks are stored in the cartridge 115, one of the optical disks is used as a metadata-dedicated optical disk, and the remaining 249 disks are used as user-data dedicated optical disks. By recording 249 user data recording positions and attribute information on one metadata-dedicated optical disc, the user-dedicated optical disc and the metadata dedicated to the user data stored in it are recorded. Can be transported and stored in a single cartridge.
  • the cartridge containing the metadata-dedicated optical disc is transported and the user-data-dedicated optical disc is stored.
  • the number of times of transportation is one, so that the access time can be shortened.
  • first access the metadata-dedicated optical disc specify the cartridge containing the user data, and then transport the cartridge containing the user-data-dedicated optical disc.
  • the procedure is to access the optical disk dedicated to user data, it is necessary to carry it twice in series, and the effect of reducing the number of times of carrying is great.
  • the metadata drive is connected to the optical disc library apparatus 101 separately from the drive for the optical disc 116 in the cartridge 115, and is different from the optical disc 116 in the cartridge 115.
  • a configuration in which an optical disk dedicated for metadata is loaded into a metadata drive and recorded is also conceivable. With this configuration, it is possible to record metadata without affecting the data access drive 114 and the optical disk 116 in the cartridge 115 and without consuming the capacity of the data optical disk. The same effect can be obtained even if the metadata drive is connected to the management server 103 or the buffering server 102. Furthermore, since there is no need to use the optical disk transport device 113, the metadata optical disk is always loaded in any state, and the time required for transporting the metadata optical disk can be reduced. is there.
  • the metadata-dedicated optical disc does not need to be one in 250.
  • the number of files is large and the ratio of the total amount of metadata is higher than the total amount of user data.
  • two of the 250 discs can be used as metadata-dedicated optical discs, and the remaining 248 discs can be used as user-data-dedicated optical discs.
  • By managing two of the plurality of drives 114 as drives for preferentially loading the metadata-dedicated optical disk all the drives 114 load the user-data-dedicated optical disk, and the metadata-dedicated optical disk is the drive 114. It is possible to avoid the situation in which the metadata cannot be accessed until the drive becomes free.
  • the optical disk 116 for recording user data often employs BD-R (TM) (Blu-ray Disc Recordable), which is an optical disk that cannot be rewritten for archiving purposes.
  • BD-RE TM
  • TM Blu-ray Disc Rewritable
  • TM Blu-ray Disc Rewritable
  • the table on the RE can be updated, and the management is simplified.
  • an optical disk since an optical disk is used, there is an advantage that it has advantages such as long life, disaster resistance, and low bit cost as in the case of non-rewritable optical disks such as BD-R.
  • FIG. 19 shows a processing flow for transition from a state managed only by the management table on the control memory in the buffering server to a state managed by the management table on the control memory in the buffering server and the management table on the optical disc. Indicates.
  • the path table 512 for each cartridge and the recording position table 513 for each file are written to the optical disc as a stub information file. (S1902) If the export condition is not met, the export process does not occur.
  • Example 5 the example shown in Example 5 will be described.
  • the pointer table 521 is updated, the cartridge ID of the cartridge that writes the path table 512 for each cartridge and the recording position table 513 for each file to the optical disk is added to the cartridge ID 1813, and the cartridge of the cartridge including the optical disk that has written the management table
  • the ID and the disk ID of the optical disk are recorded in the metadata cartridge ID 1814 and the disk ID 1815, respectively, the start position and size written on the optical disk are recorded in the start position size 1816, and the path table for each cartridge is stored in the pointer 1817. Record a pointer to 512. (S1903) It is determined whether or not the conditions for writing the pointer table 521 to the optical disc are met.
  • the management table (pointer table list 525) for managing the pointer table on the HDD is updated. That is, the cartridge ID and the disk ID of the cartridge including the written optical disk, and the written position and size are recorded in the metadata cartridge ID 1818, the disk ID 1819, and the start position size 1820 of the pointer table list 525, respectively. (S1906) Even when the writing occurs in S1902, since the pointer to the path table 512 for each cartridge is recorded in the pointer 1817, there is an effect that the management table can be read without accessing the written optical disk.
  • the path table 512 for each cartridge and the recording position table 513 for each file pointed to by the pointer 1817 are deleted, and the pointer 1817 is set to NULL. It is feasible. In this case, the management table can be read by accessing the written optical disk. Since the free memory capacity has increased, there is an effect that a larger number of cartridges can be registered in the pointer table 521.
  • FIG. 39 shows a processing flow for transition from a state managed only by the management table on the control memory in the buffering server to a state managed by the management table on the control memory in the buffering server and the management table on the optical disk. The abnormal process flow in is shown.
  • the stub information file including the logical address where the write error has occurred is registered and discarded as a non-recordable / reproducible file, the recording address is shifted, and the stub information file is recorded again in the unrecorded area. Processing can be considered. In this case, the fact that the recording address is shifted is reflected on the pointer table indicating the stub information file. Or, by using the pseudo-overwrite function (Logical Over Write) of the optical disc, the stub information file is requested to be overwritten again at the address where the error has occurred, thereby securing the necessary area in the spare area on the optical disc.
  • a process of recording the link information to the area in a DMA (Defective Management Area) of the optical disk is executed in the drive, and a process of successfully recording by retry is conceivable.
  • the management server 103 may display error information and information related to the retry processing on the management screen and notify the system administrator.
  • the pointer table 521 is updated and the conditions for writing the pointer table 521 to the optical disc are met, the pointer table 521 is written to the optical disc. If writing fails, retry processing is performed (S3904). When the number of times reaches a preset upper limit, an abnormal process is performed. (S3906) As an abnormal process, the management server 103 may display error information on the management screen and notify the system administrator. (S3906) By performing abnormality processing, the system administrator can quickly recognize the abnormality, shortening the time required to request subsequent maintenance work, and minimizing the impact on business continuity due to interruptions There is.
  • the stub information file normally recorded on the optical disk is not wasted, so that the cost of the process of re-recording the stub information file and the process of updating the pointer table becomes unnecessary. is there. Further, there is an effect that the capacity efficiency of the optical disk is not deteriorated.
  • FIG. 20 shows a processing flow for transitioning from a state managed only by the management table on the optical disk to a state managed by the management table on the control memory in the buffering server and the management table on the optical disk.
  • the determination method refers to the pointer table 520 and determines that the pointer is pointed to by the cartridge ID of the cartridge including the file and is online, provided that the pointer is not NULL.
  • the pointer table 521 is referred to, and the path table 512 for each cartridge and the recording position table 513 for each file of the cartridge including the file are read from the optical disk to the control memory in the buffering server.
  • the recording position table for each file of the file to be read does not exist in the reference destination of the pointer table 521 read on the memory, the recording position table for each file cannot be brought online.
  • the recording position table for each file exists at the reference destination of the pointer table 521, the file can be brought online.
  • the pointer to the path table 512 for each read cartridge is set in the pointer 1817 of the pointer table 521.
  • the pointer table is read out from the metadata cartridge ID 1818, the disk ID 1819, and the start position size 1820 with reference to the management table (pointer table list 525) indicating the pointer table on the optical disk.
  • the path table 512 for each cartridge and the recording position table 513 for each file of the cartridge containing the file are read from the optical disk to the control memory in the buffering server.
  • S2005 In the pointer 1817 of the read pointer table 521, a pointer to the read path table 512 for each cartridge is set.
  • the management table can be read without accessing the written optical disk.
  • FIG. 40 shows an abnormal process flow in a process flow in which the state is managed only by the management table on the optical disk to the state managed by the management table on the control memory in the buffering server and the management table on the optical disk.
  • the optical disk is transported, but if a transport error occurs (S4005), an abnormal process is performed. (S4008) If the transport is successful, the pointer table 521 is read from the optical disc. If a read error occurs (S4006), an abnormal process is performed. (S4008) If the reading process is successful, the read pointer table 521 is referenced to read the path table 512 for each cartridge and the recording position table 513 for each file from the optical disk to the control memory in the buffering server, but a read error occurs. If so (S4006), an abnormality process is performed.
  • leaving management information on the storage device has the effect of suppressing waiting time associated with cartridge transport and disk transport and ensuring accessibility and searchability.
  • the pointer table list 525 on the storage device and the pointer table 521 written to the optical disk may be transported and referenced at the maximum. The number of times can be suppressed to a maximum of 1, and the waiting time due to the multiple times of optical disk transport that is assumed when the management of this embodiment is not performed does not occur, and accessibility and searchability can be achieved even when targeting a large number of files. There is an effect that it can be secured.
  • the recording position table 513 and the pointer table are restored in disaster recovery (DR) when the buffering server 102 or the storage device 104 breaks down and becomes inaccessible due to a flood. 521 can be restored from the optical disk, and there is an effect that it is possible to avoid a situation in which the means for accessing user data is lost.
  • DR disaster recovery
  • the UDF format to the XFS format are used during DR.
  • the recording position table 513 on the optical disc and the data structure of the pointer table 521 are changed to the recording position table 513 on the storage device 104 or the like. Since the data structure of the pointer table 521 can be made the same, there is an effect that data structure conversion processing does not occur and high-speed DR is possible.
  • the recording position table 513 and the pointer table 521 are recorded on the optical disk.
  • snapshots of other management information are recorded at the same timing as the recording position table 513 and the pointer table 521 are recorded on the optical disk.
  • all management information can be recovered from the optical disk during DR, and there is no need to read out all management information distributed and recorded on the data optical disk and reconstruct the management information. , DR is effective.
  • the management table is the management information described in FIG.
  • ⁇ Process to add hash for file identification> Instead of the path table 512 for each cartridge shown in FIG. 7, a file recording position is obtained by using a hash table 522 shown in FIG. 21, a path table 512 having the same hash shown in FIG. 22, and a collision table 523 shown in FIG. Explain how to manage.
  • FIG. 21 shows a hash table 522, which includes a pointer 2112 to the same hash path table.
  • the path hash 2111 is an index shown for explanation, and is not included in the hash table 522 entity.
  • a hash operation is performed from the path, a 4-byte hash value is generated, the generated hash value is used as an index of the hash table 522, and the pointer 2112 to the path table of the same hash in the row where the hash value matches the path hash 2111 , A pointer to the same hash path table 512 is registered.
  • the row 2121 is a field used when a hash value of 0x00000000 is generated
  • the row 2122 is a field used when a hash value of 0x00000001 is generated.
  • Lines 2121 to 2123 are NULL and are unregistered.
  • a pointer is registered, which means that 0x8731b491 is generated as a hash value from the path and registered.
  • it is also registered in the row 2127 the rows 2125 through 2126 are unregistered, and the rows after 2128 are not registered.
  • the path table can be acquired by referring to the area pointed to by the pointer 0x00900230. In this way, by using the hash table, as described in S1304, it is possible to obtain a pointer to a desired table without comparing all the paths registered in the path table and confirming the duplication of the path. Has the effect of being fast.
  • FIG. 21 shows an example in which a 4-byte hash is used as an index.
  • the size of the hash is not limited to this.
  • the size of the hash table is (2 to the 32nd power row) ⁇ (size of pointer 2112 is 4 bytes), that is, 17,179,869,184 bytes, (2 8th power) ⁇ (size of pointer 2112 is 4 bytes), that is, 2,048 bytes can be reduced.
  • FIG. 22 shows the configuration of the same hash path table 512.
  • the same hash path table is a table for registering the same hash value calculated from the path, and one table is created for each hash value. Therefore, if it is a 4-byte hash value, a maximum of 2 32 tables are created.
  • the number of paths 2211 indicates the total number of file and directory paths having the same hash value generated from the path.
  • the reserve 2212 stores a fixed value and does not function in the current version, but is a reserved field for use when expanding functions in a future version.
  • the storage location 2213 of the file #N recording position table is information on the path of the Nth file or directory included in the path table 512, and a path identification hash, directory / file identification flag, recording position table pointer, collision index, which will be described later. And consists of reserves.
  • the path identification hash 2218 is a field for storing the second hash value generated from the path of the file #N.
  • the second hash value is different from the hash value generated to identify the index of the hash table 522, and is, for example, a hash value generated by another hash function.
  • the directory / file identification flag 2217 is a field for storing 0 if the file #N is a file and 1 if the file #N is a directory.
  • the recording position table pointer 2216 is a field for storing an offset value from the head in the recording position table 513 for each file.
  • the collision index 2215 indicates a value other than 0 and other than 1, and a collision table 523 described later. This is a field for registering as an index. If no collision has occurred, 0 is set. If a collision has occurred but a full path is not registered in a collision table described later, 1 is set.
  • the recording position number 2219 indicates the number of recording position tables for each file indicated by the recording position table pointer 2216.
  • the reserve 2214 stores a fixed value and does not function in the current version, but is a reserved field for use when expanding functions in a future version.
  • FIG. 23 shows a collision table 523, which consists of a full path 2312.
  • the collision index 2311 is an index shown for explanation, and is not included in the collision table 523 entity.
  • the collision index 2311 is an index from 0 to 65535. When a collision occurs in the path identification hash 2218, the collision index 2311 is allocated and used in the order of 2, 3,.
  • the collision index 0, that is, the row of 2321 is unused because it is the same as the index value 0 when no collision occurs.
  • the collision index 1, that is, the row 2322 is unused because it is the same as the index value 1 when a collision has occurred but the full path is not registered in the collision table. That is, the collision indexes 0 and 1 have special meanings and are prohibited from use for other purposes.
  • the maximum value of the collision index is 65535
  • the maximum value may be changed according to the probability of occurrence of the collision.
  • the full path 2312 is a field for recording the collided path as a full path character string. Assume a maximum of 1024 bytes.
  • the collision index 2 in the row 2323 is used, and the path where the collision has occurred is recorded in the full path 2312 with the full path.
  • FIG. 23 shows that when the hash calculation is performed on the full path “/ archive / 140101”, the same hash value as the hash value of another path that has already been registered is generated, so that a collision has occurred.
  • the full path registered in the collision table registers only the path calculated when it is confirmed that a collision has occurred, the full path that is the basis for generating the other hash value that has collided is not registered.
  • the path identification hash is calculated in order to register it in the path table 512 having the same hash, it is compared with the path identification hashes of all the files registered in the path table 512 to confirm whether or not a collision has occurred. If a collision occurs, the collision table 523 is referenced and the smallest index among the unused indexes is recorded in the collision index 2215. At the same time, the full path of the path where the collision has occurred is also recorded in the index of the collision table 523. If the collision index 2215 of the colliding file is 0, 1 is recorded. At this time, the full path is not registered in the index of the collision table 523.
  • the path identification hash matches and the collision index 2215 is 0. In this case, the search ends because it is uniquely determined at that time. If the path identification hashes match and the collision index 2215 is not 0, the full path of the corresponding index value in the collision index 2311 of the collision table 523 is referenced to check whether the file matches the full path of the file to be searched. If they match, it is the file to be searched and the search is complete.
  • search the path table 512 with the same hash using the path identification hash, and find the case where the path identification hashes match and the collision index 2215 is not 0, and check whether the full path matches in the collision table To do. This is repeated up to the same hash path table 512. If there is no matching entry, the file to be searched is found to be unregistered. Therefore, if the search is performed at the time of writing, a new storage location is stored in the same hash path table 512. What is necessary is just to add and register and to record, and if it is the search at the time of reading, the result will be that the file to search does not exist.
  • FIG. 24 is a processing flow when writing a file using a hash table.
  • FIG. 24 when writing a file designated by a certain path, two types of hash operations are performed from the designated path.
  • One is a hash value used as an index of the path hash 2111 of the hash table 522, and the other is a hash value used as the path identification hash 2218 of the path table 512 of the same hash. Assume that they are generated using a hash function a and a hash function b, respectively. (S2401) Using the hash value generated by the hash function a, the index of the path hash 2111 is specified. If the pointer 2112 is NULL, a path table 512 having the same hash is newly generated, and the pointer to the path table 512 is set. Register in the pointer 2112.
  • the same hash path table 512 pointed to by the pointer is referred to.
  • S2402 When a path table 512 having the same hash is newly generated, it is not necessary to consider whether or not a collision has occurred.
  • the path identification hash 2218 of all files (including directories) registered in the path table is compared with the hash value generated by the hash function b. Check if it has occurred.
  • S2403 If a collision has occurred, the collision table 523 is referred to and a full path is registered using an unused minimum index.
  • a new recording position is secured in the recording position table 513 for each file, and a file name (or directory name), recording time, size, and attribute are set. Further, the file number 821 (or directory number 811) is incremented. A pointer to the recording position newly secured in the recording position table 513 for each file is acquired.
  • S2405 A new storage location is secured in the same hash path table 512, and the same index value is recorded in the collision index 2215. Further, the hash value generated by the hash function b is recorded in the path identification hash.
  • the directory / file identification flag 2217 is set to 1 for a directory, and 0 for a file.
  • FIG. 41 is an abnormal process flow in the process flow at the time of writing a file using a hash table.
  • the path identification hash 2218 and the generated hash value are compared to check whether a collision has occurred.
  • the collision table 523 is referred to and it is confirmed whether an unused minimum index can be used.
  • S4102 If the number of registrations reaches the upper limit and there is no unused index, abnormal processing is performed.
  • S4106 If no collision has occurred in S4101, and if registration has succeeded in S4102, it is confirmed whether registration is possible in the recording position table 513 for each file.
  • S4103 If the number of registrations reaches the upper limit and registration fails, abnormal processing is performed.
  • FIG. 25 is a processing flow when reading a file using a hash table.
  • FIG. 25 when reading a file specified by a certain path, two types of hash operations are performed from the specified path.
  • One is a hash value used as an index of the path hash 2111 of the hash table 522, and the other is a hash value used for searching the path identification hash 2218 of the path table 512 of the same hash.
  • it is necessary to generate them using the hash function a and the hash function b, respectively. (S2501) If the index of the path hash 2111 is specified using the hash value generated by the hash function a and the pointer 2112 is NULL, the specified file does not exist.
  • the same hash path table 512 pointed to by the pointer is referred to.
  • the path identification hash 2218 of all files (including directories) registered in the path table 512 is compared with the hash value generated by the hash function b to confirm whether a collision has occurred.
  • the index value of the collision index 2215 is acquired, and if it is 1, it is skipped and the next collision is searched. If the value is other than that, the full path 2312 is acquired by referring to the row of the index in the collision table 523. Compare the full path with the full path of the specified file and check if they match.
  • the recording position table pointer 2216 of the storage location having the index of the same hash path table 512 is acquired. If they do not match, return to the hash table 512 of the same path and continue the search, compare the path identification hash 2218 and the hash value generated by the hash function b, identify the location where the next collision occurs, Repeating the full path comparison at the collision table 523 is repeated. If there is no match, a pointer to the recording position table whose collision index is 1 is acquired. (S2505) Follow the acquired pointer to acquire the recording position and read the file.
  • FIG. 42 is an abnormal process flow in the process flow when reading a file using a hash table.
  • the index of the path hash 2111 is specified using the hash value generated by the hash function a and the pointer 2112 is NULL, the designated file does not exist, and the pointer to the path table is Since it cannot be acquired (S4201), an abnormality process is performed.
  • the path identification hash 2218 of all files (including directories) registered in the path table 512 and the hash value generated by the hash function b are compared to confirm whether a collision has occurred.
  • the full path 2312 is acquired by referring to the row of the index in the collision table 523. Compare the full path with the full path of the specified file and check if they match.
  • the size of the full path 2312 can be reduced from 1024 bytes. For example, if a hash value with a length of 40 bytes, which is the same size as the UUID, is generated, it is sufficient because the collision probability is low enough to be regarded as no collision occurs.
  • FIG. 26 shows the configuration of the same hash path table 512.
  • the description of the same part as the same hash path table 512 shown in FIG. 22 will be omitted, and only the differences will be described.
  • the drive ID + time stamp 2617 is an ID obtained by combining a unique serial number assigned for identifying the drive 114 mounted on the optical disc library apparatus 101 and a time stamp when the file is recorded on the optical disc. For example, it is conceivable that the upper 8 bytes are the serial number of the drive 114 and the lower 8 bytes are a time stamp, giving an ID of 16 bytes in total.
  • FIG. 27 is a processing flow when writing a file using a time stamp in the ID.
  • a hash operation is performed from the designated path.
  • This is a hash value used as an index of the path hash 2111 of the hash table 522.
  • S2701 Using the generated hash value, the index of the path hash 2111 is specified. If the pointer 2112 is NULL, a path table 512 having the same hash is newly generated, and the pointer to the path table 512 is registered in the pointer 2112. To do. If not NULL, the same hash path table 512 pointed to by the pointer is referred to.
  • S2702 Record the file and get the time stamp when it was recorded. Also, the serial number of the recorded drive is acquired.
  • a new recording position is secured in the same hash path table 512, and is set to drive ID + time stamp 2617.
  • S2703 A new recording position is secured in the recording position table 513 for each file, and a file name (or directory name), recording time, size, and attribute are set. Further, the file number 821 (or directory number 811) is incremented. A pointer to the recording position newly secured in the recording position table 513 for each file is acquired.
  • S2704 A new storage location is secured in the same hash path table 512, and the directory / file identification flag 2616 is set to 1 for a directory and set to 0 for a file.
  • FIG. 43 is an abnormal process flow in the process flow at the time of writing a file using a time stamp as an ID.
  • a hash operation is performed from the specified path, the index of the path hash 2111 is specified using the generated hash value, and the pointer 2112 is NULL. If there is, a new path table 512 with the same hash is generated and a pointer to the path table 512 is registered in the pointer 2112 (S4301). If registration fails, an abnormal process is performed. (S4305) If it is not NULL, the path table 512 of the same hash pointed to by the pointer is referred to, and the pointer to the path table 512 is registered in the pointer 2112 (S4301). If registration fails, an abnormal process is performed.
  • S4305 A pointer to the recording position newly secured in the recording position table 513 for each file is acquired, a storage location is newly secured in the path table 512 of the same hash, the directory / file identification flag 2616 is set, and the recording position table pointer In 2615, a pointer to a newly secured recording position is set in the recording position table 513 for each file. If the pointer fails, an abnormal process is performed. (S4305) If it is completed normally, the system administrator is notified (S4304). As an abnormal process, the management server 103 may display error information on the management screen and notify the system administrator. (S4305) By performing abnormality processing, the system administrator can quickly recognize the abnormality, shortening the time required to request subsequent maintenance work, and minimizing the impact on business continuity due to interruptions There is.
  • FIG. 28 is a processing flow when reading a file using a time stamp in the ID.
  • a hash operation is performed from the specified path.
  • This is a hash value used as an index of the path hash 2111 of the hash table 522.
  • S2801 Using this hash value, the index of the path hash 2111 is specified, and if the pointer 2112 is NULL, the designated file does not exist. If not NULL, the same hash path table 512 pointed to by the pointer is referred to.
  • S2802 The ID specified together with the path is compared with the drive ID + time stamp 2617 of the path table 512 having the same hash, and the matching ID is specified. (S2803) If they do not match, the specified file does not exist.
  • FIG. 44 is an abnormality processing flow in the processing flow at the time of reading a file using a time stamp as an ID.
  • a hash operation is performed from the specified path, and if the pointer 2112 is NULL (S4400), the specified file does not exist, so an abnormal process is performed. I do. (S4404) If it is not NULL, the hash value is used to identify the index of the path hash 2111, refer to the same hash path table 512 pointed to by the pointer, the ID specified together with the path, and the drive ID of the same hash path table 512+ The time stamps 2617 are compared to identify matching ones, but if they do not match (S4401), the designated file does not exist and abnormal processing is performed.
  • the recording position table pointer 2615 is acquired, and the recording position is acquired by following the acquired pointer (S4402).
  • S4404 If it is completed normally, the system administrator is notified (S4403).
  • the management server 103 may display error information on the management screen and notify the system administrator. (S4404) By performing abnormality processing, the system administrator can quickly recognize the abnormality, shortening the time required to request subsequent maintenance work, and minimizing the impact on business continuity due to interruptions There is.
  • the use of the time stamp for the ID has an effect that the file can be identified without using a process of performing a hash operation for a large number of files many times.
  • FIG. 31 is a diagram showing the structure of the owner ID used for the owner ID 1011 of the share information table 515 and the owner ID 1112 of the cartridge information table 516.
  • 24 to 31 bytes (3111) is a portion corresponding to the upper 4 bytes of the owner ID having a length of 32 bytes, and is a unique ID such as a serial number of the buffering server 102.
  • 16 to 23 bytes (3112) is a portion corresponding to 16 to 23 bytes of the owner ID, and indicates the date when the ID was generated in BCD.
  • 8-15 bytes (3113) is a portion corresponding to 8-15 bytes of the owner ID, and indicates the time (hour minute second) when the ID is generated in BCD.
  • the description has been made with the resolution up to the hour, minute, and second, it is possible to maintain a smaller resolution such as milliseconds or microseconds by changing the bit assignment.
  • 0 to 7 bytes (3114) is a portion corresponding to the lower 4 bytes of the owner ID, and is a value generated by random number calculation.
  • FIG. 29 is a processing flow for performing file writing using the owner ID shown in FIG.
  • the serial number and time stamp of the buffering server 102 are acquired, and the owner ID is generated.
  • the file is recorded using the generated owner ID as part of the path.
  • S2902 Write processing as shown in the seventh embodiment is performed.
  • S2903 The latest recording time 613 of the common information 511 for each cartridge of the cartridge in which the file is recorded is updated.
  • FIG. 45 is an abnormal process flow in the process flow for writing a file using the owner ID shown in FIG.
  • the serial number of the buffering server 102 is acquired (S4501), but if it fails, abnormal processing is performed. (S4505) If the time stamp is acquired (S4502) but fails, abnormal processing is performed. (S4506) An owner ID is generated from the serial number and the time stamp. If the generated owner ID is duplicated (S4503), an abnormality process is performed. (S4505) A file is recorded using the generated owner ID as a part of the path (S4504), but if it fails, abnormal processing is performed. (S4506) Write processing as shown in the seventh embodiment is performed. (S2903) If it is completed normally, the system administrator is notified (S4505). As an abnormal process, the management server 103 may display error information on the management screen and notify the system administrator. (S4506) By performing abnormality processing, the system administrator can quickly recognize the abnormality, shortening the time required to request subsequent maintenance work, and minimizing the impact on business continuity due to interruptions There is.
  • FIG. 30 is a processing flow for reading a file using the owner ID shown in FIG.
  • the owner ID is extracted from the path.
  • a time stamp (year / month / day, hour / minute / second) is extracted from the owner ID.
  • S3002 By referring to the common information 511 for each cartridge from the pointer table 521, the recording time range for each cartridge can be acquired from the oldest recording time 612 and the latest recording time 613 recorded in the common information 511, and the recording time range is set in S3002. The cartridge ID of the cartridge including the obtained time stamp can be acquired.
  • FIG. 46 is an abnormal process flow in the process flow of reading a file using the owner ID shown in FIG.
  • the owner ID is extracted from the path (S4601), but if it fails, abnormal processing is performed.
  • S4605 A time stamp (year / month / day, hour / minute / second) is extracted from the owner ID (S4602), but if it fails, abnormal processing is performed.
  • S4605 By referring to the common information 511 for each cartridge from the pointer table 521, the recording time range for each cartridge is acquired from the oldest recording time 612 and the latest recording time 613 recorded in the common information 511 (S4603). If it does, abnormal processing is performed.
  • FIG. 49 is a processing flow for writing a file using the owner ID shown in FIG. 31 and the update processing of the share information table shown in FIG.
  • the serial number and time stamp of the buffering server 102 are acquired, and the owner ID shown in FIG. 31 is generated.
  • the file is recorded using the generated owner ID as part of the path.
  • the path has a configuration of, for example, / OwnerID / DirectoryName / FileName.
  • OwnerID is a character string of the share name
  • the owner ID 1011 can be acquired from the share 1012 from the share information table 515.
  • DirectoryName indicates a character string of the name of the directory storing the File
  • “FileName” indicates a character string of the name of the file to be recorded.
  • “/” Is a character indicating a delimiter.
  • FIG. 47 is a processing flow showing the update processing of the share information table.
  • the share information table 515 is written as a share information file on the optical disc. (S4702) If the export condition is not met, the export process does not occur.
  • the pointer table 521 is updated, the cartridge ID of the cartridge that writes the share information table 515 to the optical disk is added to the cartridge ID 1813, and the cartridge ID of the cartridge that includes the optical disk that has written the management table and the disk ID of the optical disk are The start position and size written in the metadata cartridge ID 1814 and the disk ID 1815 and written on the optical disk are recorded in the start position size 1816, and the pointer to the path table 512 for each cartridge is recorded in the pointer 1817. Further, the minimum owner ID and the maximum owner ID among the owner IDs of the cartridge share of the pointer table 521 are held. (S4703) Next, it is determined whether or not the conditions for writing the pointer table 521 to the optical disc are met. (S4704) As a condition, it may be considered that a specified period of time has passed since the last update time of the share information table. A case where the number of lines recorded in the pointer table 521 is equal to or greater than a threshold value can be considered.
  • the pointer table 521 is written to the optical disc. Further, the management table (pointer table list 525) for managing the pointer table on the HDD is updated. That is, the cartridge ID and the disk ID of the cartridge including the written optical disk, the written position and size, the retained minimum owner ID and the maximum owner ID are respectively stored in the metadata cartridge ID 1818 of the pointer table list 525 and the disk. Recorded in ID 1819, start position size 1820, and minimum OID and maximum OID 1821. (S4705) If the export condition is not met, the export process does not occur.
  • the management table can be read by accessing the written optical disk. Since the free memory capacity has increased, there is an effect that a larger number of cartridges can be registered in the pointer table 521.
  • FIG. 50 is a process flow for reading a file using the owner ID shown in FIG. 31 and the share information table acquisition process shown in FIG.
  • the owner ID is extracted from the path.
  • the path has a configuration of, for example, / OwnerID / DirectoryName / FileName.
  • OwnerID is a character string of the share name
  • the owner ID 1011 can be acquired from the share 1012 from the share information table 515.
  • “DirectoryName” indicates a character string of the name of the directory storing the File
  • “FileName” indicates a character string of the name of the file to be recorded.
  • “/” Is a character indicating a delimiter. Although only one “DirectoryName” is described in this example, it may have a hierarchical structure.
  • a time stamp (year / month / day, hour / minute / second) is extracted from the owner ID.
  • a time stamp (year / month / day, hour / minute / second) is extracted from the owner ID.
  • a portion corresponding to 16 to 23 bytes 3112 is extracted and converted from BCD to a numerical value.
  • a portion corresponding to 8 to 15 bytes 3113 may be extracted and converted from BCD to a numerical value.
  • S5002 a share information table acquisition process as shown in FIG. 48 is performed.
  • S5003 Next, the share information table 515 acquired by the share information acquisition process of FIG.
  • FIG. 48 is referred to, the owner ID 1011 is acquired from the share name 1012 included in the path, and the cartridge information table 516 is further referred to, and the cartridge is determined from the owner ID 1112. ID1111 is acquired.
  • S5004 Since the metadata cartridge and the optical disk holding the management information of the cartridge with the acquired cartridge ID can be seen from the pointer table 521, the optical disk is loaded from the metadata cartridge, and the stub information file is read out, so that it is shown in the seventh embodiment. Read processing is performed.
  • FIG. 48 is a process flow showing a share information table acquisition process.
  • the determination method refers to the pointer table 520 and determines that the pointer is pointed to by the cartridge ID of the cartridge including the file and is online, provided that the pointer is not NULL.
  • the pointer table 521 is referred to, and the share information table 515 of the cartridge including the file is read from the optical disk to the control memory in the buffering server.
  • the share information table of the file to be read does not exist at the reference destination of the pointer table 521 read onto the memory, the share information table cannot be brought online. On the other hand, when the share information table exists in the reference destination of the pointer table 521, it can be brought online.
  • the pointer to the read share information table is set in the pointer 1817 of the pointer table 521.
  • the management table pointer table list 525) pointing to the pointer table on the optical disk is referenced, the cartridge and optical disk whose owner ID is included in the minimum OID maximum OID 1821 are specified, and the optical disk is transported Read the pointer table.
  • the share information table 515 of the cartridge including the file is read from the optical disk to the control memory in the buffering server. Which of the share information tables indicated by the pointer table 521 corresponds to the share information table of the desired cartridge can be determined by narrowing the range of the minimum OID and maximum OID 1821 by binary tree search using the time stamp included in the owner ID. Can be searched efficiently.
  • the share information table 515 pointed to by the pointer 1817 is preferentially deleted and the pointer 1817 is set to NULL. It can be realized by setting.
  • the pointer table list 525 on the storage device and the pointer table 521 written to the optical disc may be transported and referenced at the maximum.
  • the number of times can be suppressed to a maximum of 1, and the waiting time due to the multiple times of optical disk transport that is assumed when the management of this embodiment is not performed does not occur, and accessibility and searchability can be achieved even when targeting a large number of files. There is an effect that it can be secured.
  • the present invention is not limited to the above-described embodiments, and includes various modifications.
  • the above-described embodiments have been described in detail for easy understanding of the present invention, and are not necessarily limited to those having all the configurations described.
  • a part of the configuration of one embodiment can be replaced with the configuration of another embodiment, and the configuration of another embodiment can be added to the configuration of one embodiment.
  • Each of the above-described configurations, functions, processing units, processing means, and the like may be realized by hardware by designing a part or all of them with, for example, an integrated circuit.
  • Each of the above-described configurations, functions, and the like may be realized by software by interpreting and executing a program that realizes each function by the processor.
  • the program for realizing each function can be placed on a recording device such as a hard disk or SSD (Solid State Drive), or a recording medium such as an IC card, SD card, or DVD.
  • the control lines and information lines indicate what is considered necessary for the explanation, and not all the control lines and information lines on the product are necessarily shown. Actually, it may be considered that almost all the components are connected to each other.
  • Optical Disk Library Device 101 Optical Disk Library Device 102 Buffering Server 103 Management Server 104 Storage Device 105 Host 111 CPU 112 Control Memory 113 Optical Disk Transport Device 114 Drive 115 Cartridge 116 Optical Disk 211 CPU 212 Control Memory 311 Server Command Control Unit 312 Resource Management Unit 411 Host Command Control Unit 412 Recording Position Management Unit 413 Migration Control Unit

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Management Or Editing Of Information On Record Carriers (AREA)

Abstract

L'invention concerne un système de mémorisation qui peut réduire le transport de disques optiques et garantir des performances d'accès tout en limitant la capacité et les coûts de dispositifs de mémorisation HDD et en diminuant les coûts de migration de données de gestion sur un HDD, grâce : à l'écriture, à un moment approprié, sur un disque précis dans un certain dispositif de mémorisation HDD à cartouche de disques optiques, d'informations de gestion de position d'enregistrement indiquant quels fichiers sont mémorisés dans quels disques optiques à l'intérieur d'une cartouche, dans un cas où des opérations sont en cours pour stocker des cartouches externes à un dispositif bibliothèque de disques optiques ; et à la lecture des informations de gestion de position d'enregistrement à partir d'un disque optique en fonction d'une demande de lecture de fichier provenant d'un utilisateur.
PCT/JP2015/066647 2015-06-10 2015-06-10 Système de mémorisation WO2016199230A1 (fr)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/JP2015/066647 WO2016199230A1 (fr) 2015-06-10 2015-06-10 Système de mémorisation
JP2017522797A JPWO2016199230A1 (ja) 2015-06-10 2015-06-10 ストレージシステム

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/JP2015/066647 WO2016199230A1 (fr) 2015-06-10 2015-06-10 Système de mémorisation

Publications (1)

Publication Number Publication Date
WO2016199230A1 true WO2016199230A1 (fr) 2016-12-15

Family

ID=57504759

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2015/066647 WO2016199230A1 (fr) 2015-06-10 2015-06-10 Système de mémorisation

Country Status (2)

Country Link
JP (1) JPWO2016199230A1 (fr)
WO (1) WO2016199230A1 (fr)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPWO2021171683A1 (fr) * 2020-02-28 2021-09-02
CN117271440A (zh) * 2023-11-21 2023-12-22 深圳市云希谷科技有限公司 一种基于freeRTOS文件信息存储方法、读取方法及相关设备

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2011253588A (ja) * 2010-06-02 2011-12-15 Hitachi-Lg Data Storage Inc 光ディスク装置及び光ディスクの記録再生方法
JP2015088207A (ja) * 2013-10-31 2015-05-07 株式会社日立エルジーデータストレージ ライブラリ装置及びこれを有するアーカイブ装置

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2011253588A (ja) * 2010-06-02 2011-12-15 Hitachi-Lg Data Storage Inc 光ディスク装置及び光ディスクの記録再生方法
JP2015088207A (ja) * 2013-10-31 2015-05-07 株式会社日立エルジーデータストレージ ライブラリ装置及びこれを有するアーカイブ装置

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPWO2021171683A1 (fr) * 2020-02-28 2021-09-02
WO2021171683A1 (fr) * 2020-02-28 2021-09-02 富士フイルム株式会社 Dispositif de traitement d'informations, procédé de traitement d'informations et programme de traitement d'informations
EP4113269A4 (fr) * 2020-02-28 2023-08-09 FUJIFILM Corporation Dispositif de traitement d'informations, procédé de traitement d'informations et programme de traitement d'informations
US11763840B2 (en) 2020-02-28 2023-09-19 Fujifilm Corporation Information processing device, information processing method, and information processing program
CN117271440A (zh) * 2023-11-21 2023-12-22 深圳市云希谷科技有限公司 一种基于freeRTOS文件信息存储方法、读取方法及相关设备
CN117271440B (zh) * 2023-11-21 2024-02-06 深圳市云希谷科技有限公司 一种基于freeRTOS文件信息存储方法、读取方法及相关设备

Also Published As

Publication number Publication date
JPWO2016199230A1 (ja) 2017-07-27

Similar Documents

Publication Publication Date Title
US10977124B2 (en) Distributed storage system, data storage method, and software program
EP3617867B1 (fr) Procédé de gestion de fragment et appareil de gestion de fragment
US7933938B2 (en) File storage system, file storing method and file searching method therein
US10365983B1 (en) Repairing raid systems at per-stripe granularity
US7827146B1 (en) Storage system
US20110010496A1 (en) Method for management of data objects
US10503424B2 (en) Storage system
CN106445405B (zh) 一种面向闪存存储的数据访问方法及其装置
US7941406B2 (en) Techniques for snapshotting
US20070061540A1 (en) Data storage system using segmentable virtual volumes
JP2016509310A (ja) 格納されたデータにおける冗長削減
KR101356470B1 (ko) 플래시 파일 시스템
US8140886B2 (en) Apparatus, system, and method for virtual storage access method volume data set recovery
US9934248B2 (en) Computer system and data management method
US7895164B1 (en) Efficient checkpoint process
CN106095331B (zh) 一种固定大文件内部资源的控制方法
CN108021562B (zh) 应用于分布式文件系统的存盘方法、装置及分布式文件系统
WO2016199230A1 (fr) Système de mémorisation
US8880794B2 (en) Populating a sparsely allocated memory in a storage array
JP4828917B2 (ja) ストレージシステム、計算機システム、ホスト計算機及びデータ複製方法
CN116431062A (zh) 数据存储方法、设备、存储介质和系统
US7437528B1 (en) Gang blocks
JP4390618B2 (ja) データベース再編成プログラム、データベース再編成方法、及びデータベース再編成装置
CN111625477B (zh) 访问擦除块的读请求的处理方法与装置
CN105573862B (zh) 一种恢复文件系统的方法和设备

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 15894915

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2017522797

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 15894915

Country of ref document: EP

Kind code of ref document: A1