JP3750163B2 - Cache disk control - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a cache disk for hierarchical storage management control using a magnetic disk or RAID as a primary storage device and using an optical disk, a compact disk, or a magnetic tape as a secondary storage device.
[0002]
[Prior art]
As a method of moving user data written from a primary storage device to a secondary storage device (optical disk, compact disk, magnetic disk, etc.) with a pair of primary storage devices for each secondary storage unit, the following two methods There is.
[0003]
1) Designate a time when the user is not using it and move it at that time.
[0004]
2) Basically, 1), but when the write capacity of the primary storage device is full, the movement of write data is started.
[0005]
[Problems to be solved by the invention]
It is avoided that the user data write capacity of the primary storage device becomes full during the user use time period, and the data write performance for the user is deteriorated.
[0006]
[Means for Solving the Problems]
In addition to the primary storage device corresponding to each secondary storage device structural unit, the entire secondary storage device includes a primary storage device that can be used in common. When the user data write capacity of the primary storage device corresponding to the secondary storage structural unit becomes full during the user usage time period, the user data temporarily written in the common available primary storage device described above And the user data is moved to a secondary storage device (optical disk, compact disk, magnetic tape, etc.) during a specific time zone not used by the user. For this reason, the primary storage device does not become full during the user use time zone, and the data writing performance for the user does not deteriorate.
[0007]
DETAILED DESCRIPTION OF THE INVENTION
What has been described above will be described using the configuration shown in FIG.
[0008]
1) Configuration i) One server workstation
ii) Optical disk library device (secondary storage device)
A total of 4 volumes with 2 optical disks as 1 volume
iii) Magnetic disk (primary storage device)
Partition 1-4 shared use cache partition 5 as cache for each volume of optical disk library device
iv) User client 4 sets of user data are registered in the server workstation using NFS. v) Hierarchical storage management directory configuration The optical disk volume 1 and the cache partition 1 constitute a directory / L1.
[0009]
Similarly, / L2, / L3, / L4 are configured.
[0010]
2) It is assumed that data is registered in the hierarchical storage management directory / L1- / L4 of the server from the operating user client.
[0011]
(1) A control method in the case where data registration occurs exceeding the data writing capacity of the cache partition and the data is temporarily saved in the shared cache partition.
[0012]
As shown in FIG. 2, since the data writing area of the cache partition 1 is full, a part of the data is temporarily moved to the cache partition 5. At this time, the attribute information of the moved data (file) is held in the data management area of the cache partition 1, and only the substance of the data (file) is moved to the shared use cache partition 5. At this time, the physical location information of the data moved to the cache partition 5 is written in the data management area of the cache partition 1 and a flag indicating that a part of the data is in the cache partition 5 is set. In the data management area of the cache partition 5, physical location information and use area capacity occupied by the data of the cache partition 1 are registered, and it is the cache partition 1 that uses the area. Also register the information.
[0013]
(2) A control method in the case where a read request is issued from the user client to the data saved in the shared use cache partition.
[0014]
Reading is performed using the management data registered at the time of the control (1).
[0015]
First, the physical position of the data on the cache partition 5 registered in the data management area of the cache partition 1 is obtained, and the target data is transferred from the cache partition 5 to the user client.
[0016]
(3) A control method when a data transfer request for data saved on the shared use cache partition to the optical disk is issued from the server workstation.
[0017]
This requests the cache partition 1 to first transfer data to the optical disk. At that time, if there is a flag indicating that there is data at the location of the cache partition 5 in the data management area of the cache partition 1, the physical location information of the data in the cache partition 5 is also registered as shown in (1). Therefore, data transfer is performed from the cache partition 5 to the optical disc from the position information. After the data in the cache partition 5 has been transferred to the optical disk, the data in that area is no longer needed, so the flag indicating that the cache partition 1 was used from the data management area of the cache partition is used. Delete physical location information and used capacity data.
[0018]
【The invention's effect】
Even if the user data writing capacity of the primary storage device becomes full during the user usage time, the data writing performance for the user does not deteriorate.
[Brief description of the drawings]
FIG. 1 is a conceptual diagram for explaining a configuration of the present invention.
FIG. 2 is a conceptual diagram of a cache memory for explaining the configuration of the present invention.
[Explanation of symbols]
1, 2, 3, 4 ... Cache partition corresponding to optical disk volume, 5 ... Shared cache partition.

Claims (1)

A first storage device that is divided into a plurality of areas and that writes data based on a user data write request; and a data access speed that is slower than that of the first storage device; A hierarchical storage system comprising a plurality of second storage devices associated with the plurality of areas and writing data written in the first storage device to the second storage device at a user-specified time There,
The first storage device is configured to write user data when user data is written exceeding the capacity of any storage area associated with the second storage device. Have other storage areas that are not supported,
The hierarchical storage system, wherein the storage area associated with the second storage device has a management information area for recording position information of user data written to the other area.

Images