Storage system has the function of preventing drive write error
AssigneeHitachi, Ltd.
InventorsYutaka Nakagawa, Ikuya Yagisawa, Akira Nishimoto
Life of this patent
Estimating the term…
Abstract
A storage technology for improving reliability in writing host data to a storage device is provided. A first check code based on write data is stored in cache memory 370, and storage section 50 is instructed to write, in specified predetermined storage area, write data with the first check code appended thereto, after which the storage section 50 is instructed to read data stored in said predetermined storage area; a second check code is generated on the basis of the read data, and in the event that the correspondence relationship between the first and second check codes is correct, it is decided that the write data was written normally to the storage section 50.
Claims11
A storage system for saving host data handled by a host computer, comprising:
- a plurality of storage devices having a plurality of storage areas capable of storing data;
- a RAID controller that controls said plurality of storage devices by RAID technology,
- a write data receiving unit that receives host data instructed to be saved from said host computer as write data;
- a check code appending unit that generates a first check code having a relationship with said write data, and appends said first check code to said write data;
- a cache memory that stores said first check code;
- a write instructing unit that instructs said storage device on a write operation for writing said write data having said first check code, to a specified storage area among said plurality of storage areas;
- a read instructing unit that, after said write operation, instructs said storage device to read data stored in said specified storage area as read data;
- a check code generating unit that generates a second check code having a relationship with said read data;
- a writing check unit that checks a correspondence between said first check code stored in said cache memory and said generated second check code in order to determine whether said write operation has been done normally;
- a recovering unit that recovers data saved on a storage device for which error is detected by said writing check unit;
- and
- an overwriting unit that overwrites said recovered data onto said storage device for which error is detected.
A data writing method in a storage system that saves host data handled by a host computer, comprising the steps of:
- providing a plurality of storage devices that have a plurality of storage areas, and a RAID controller that controls writing of said host data to said storage devices,
- receiving host data instructed to be saved from said host computer as write data;
- generating a first check code having a relationship with said write data, and appending said first check code to said write data;
- storing said first check code in a cache memory, by said controller;
- instructing said storage device in a write operation to write said write data having said first check code, to a specified storage area among said plurality of storage areas;
- after said write operation, instructing said storage device to read data stored in said specified storage area as read data;
- generating a second check code having a relationship with said read data;
- checking a correspondence between said first check code stored in said cache memory and said generated second check code in order to determine whether said write operation has been done normally;
- recovering data saved on a storage device for which error is detected in said step of checking correspondence;
- and
- overwriting recovered data onto said storage device for which error is detected.
A data writing method in a storage system that saves host data handled by a host computer, comprising the steps of:
- providing a plurality of storage device that have a plurality of storage areas, and a RAID controller that controls writing of said host data to said storage devices,
- receiving host data instructed to be saved from said host computer as write data;
- generating a first check code having a relationship with said write data, and appending said first check code to said write data;
- storing said first check code in a cache memory, by said controller;
- instructing said storage device in a write operation write said write data having said first check code, to a specified storage area among said plurality of storage areas;
- after said write operation, instructing said storage device to read data stored in said specified storage area as read data;
- extracting a check code from said read data as a second check code;
- checking a correspondence between said first check code stored in said cache memory and said extracted second check code in order to determine whether said write operation has been done normally;
- recovering data saved on a storage device for which error is detected in said step of checking correspondence;
- and
- overwriting recovered data onto said storage device for which error is detected.
Dependencies are inferred from the claim text.