Data access management system in distributed processing system

• US 7,228,352 B1
• Filed: 10/10/2000
• Issued: 06/05/2007
• Est. Priority Date: 12/27/1999
• Status: Expired due to Fees

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Yaguchi, Toshihiko, Yokoyama, Mamoru
• Primary Examiner(s): Wiley; David
• Assistant Examiner(s): Neurauter, Jr.; George
• Application Number:
  US09/685,615
• Time in Patent Office: 2,429 Days
• Field of Search: 709201-253, 714/15, 714/19
• US Class Current: 709/229
• CPC Class Codes: G06F 11/1662 the resynchronized component or unit being a persistent storage device re-synchronization of failed mirror storage G06F11/2082; rebuild or reconstruction of parity RAID storage G06F11/1008
• CPC Class Codes: G06F 11/2028 eliminating a faulty processor or activating a spare
• CPC Class Codes: G06F 11/2033 switching over of hardware resources
• CPC Class Codes: G06F 11/2041 with more than one idle spare processing component
• CPC Class Codes: G06F 11/2046 where the redundant components share persistent storage G06F11/2043 takes precedence
• CPC Class Codes: G06F 12/1458 by checking the subject access rights
• CPC Class Codes: G06F 3/0601 Interfaces specially adapted for storage systems
• CPC Class Codes: G06F 3/0673 Single storage device

Abstract

A data access management system in a distributed processing system capable of immediately restoring the system as a whole, has computers connected to each other via LAN, and a secondary storage device to which the computers are connected via data transfer paths. An access control subsystem for executing a data access to the secondary storage device is executed on each of the computers. A file management subsystem for giving an authority for writing the data to each access control subsystem, is executed on specified computer. When the file management subsystem transmits, to a certain access control subsystem, access control information indicating the authority for writing the data to a specified block on the secondary storage device, the same access control information is saved in the transmitted-side access control subsystem, and its record is retained in the file management subsystem. If the file management subsystem falls into a process-down, a standby file management subsystem is determined to become a new file management subsystem. Then this standby file management subsystem requests each of the access control subsystems to notify of the access control information saved by each of the access control subsystems.

Claims

1. A data access management system in a distributed processing system comprising:

a plurality of computers connected to each other under a network environment; and

a shared disk,

wherein each of said computers has an access control subsystem capable of retaining a write-only token which specifies an area in the shared disk and represents authority to exclusively write data onto a specified area in the shared disk, and a read-only token which specifies an area in the shared disk and represents authority to non-exclusively read data from a specified area in said shared disk, said access control subsystem writes, when retaining a write-only token, data on the area specified by the write-only token, and said access control subsystem reads, when retaining a read-only token, data from the area specified by the read-only token,

wherein any one of said computers has a management subsystem for uniquely generating a write-only token for every area in the shared disk, respectively issuing the generated write-only tokens for an access control subsystem requesting a write-only token, and generating a read-only token for an area for which no write-only token has been issued and issuing the generated read-only token to an access control subsystem requesting a read-only token, said management subsystem recording which access control subsystem is issued with a write-only token and which area is specified by the same write-only token, and

wherein at least one of said computers has a standby management subsystem for eliciting, if said management subsystem is incapable of continuing the process, from each of said access control subsystems whether or not it retains a write-only token and which area is specified by its retaining write-only token, generating a write-only token for an area which is specified by no write-only token responded from any access control subsystem and recording which access control subsystem is issued with a write-only token and which area is specified by the same write-only token based on a response to the elicitation from an access control subsystem.

2. A data access management system in a distributed processing system according to claim 1, wherein said management subsystem sets, for said access control subsystem, a part of the storage area of said shared disk as a reserve area that can be allocated to data by said access control subsystem, and said setting of said reserve area includes issuing said write-only token for all blocks in said reserve area.

3. A data access management system in a distributed processing system according to claim 2, wherein said access control subsystem records in an area allocated to the data within the reserve area set by said management subsystem as an on-use area, and notifies said management subsystem of information about the on-use area,

wherein said management subsystem records the information about the on-use area of which said access control subsystem has notified, and

wherein said standby management subsystem, if said management subsystem is incapable of continuing the process, requests each of said access control subsystem for the information about the on-use area that is recorded by said access-control subsystem, and records the information about the on-use area of which said access control subsystem has notified in response to this request.

4. A data access management system in a distributed processing system according to claim 3, wherein said shared disk retains management information indicating, for every block in the storage area, that it is set as reserve area for any one of said access control subsystems, that it is an on-use area allocated to data by any one of said access control subsystems or that it is a free area, and log data recorded with a history of updating the management data,

wherein said management subsystem, each time it sets the reserve area and each time it is notified of the information about the on-use area, records the facts in the log data, and

wherein said standby management subsystem, if said management subsystem is incapable of continuing the process, updates the management information on the basis of the log data.

5. A data access management system in a distributed processing system according to claim 4, wherein said standby management subsystem, if said management subsystem is incapable of continuing the process, makes each of said access control subsystems execute a redundant transmission of a procedure transmitted by said access control subsystem to said management subsystem.

6. A data access management system in a distributed processing system according to claim 5, wherein said standby management subsystem, if the log data contains a content corresponding to the procedure redundantly transmitted from said access control subsystem, replies that this procedure succeeds to said access control subsystem, and, if the log data does not contain the content corresponding to the procedure redundantly transmitted from said access control subsystem, replies that this procedure falls into a failure to said access control subsystem.

7. A data access management system in a distributed processing system according to claim 1, wherein any one of said computers includes a monitor subsystem for monitoring an operation of said management subsystem and, when detecting that said management subsystem is incapable of continuing the process, operating said standby management subsystem.

8. A data access management system in a distributed processing system according to claim 7, wherein each of said computers includes said standby management subsystem,

said monitor subsystem, when detecting that said management subsystem is incapable of continuing the process, operates only one of said standby management subsystems.