Disk area isolation method and device

Patent metadata

Field	Value
Publication number	US-10585592-B2
Application number	US-201715445727-A
Country	US
Kind code	B2
Filing date	Feb 28, 2017
Priority date	Aug 30, 2014
Publication date	Mar 10, 2020
Grant date	Mar 10, 2020

Abstract

Embodiments of the present disclosure provide a disk area isolation method, where the method includes: receiving, by a file system, a request for access to a file, where the access includes file reading or file writing; determining a to-be-accessed data block according to the request; accessing a physical area corresponding to the data block on a disk; identifying, according to a time consumed for accessing the physical area, whether the data block is located in a low-speed area on the disk; and isolating the data block that is identified to be located in the low-speed area.

First claim

What is claimed is:

1. A disk area isolation method, wherein the method is applied to a file system, the file system stores a file in a unit of data block, each data block has a corresponding physical area on a disk, and the method comprises:
receiving, by the file system, a request for access to a file, wherein the access comprises file reading or file writing, wherein the file system comprises a data structure that is used to describe logical partitioning of the disk, and the data structure comprises an isolation block and a data area, wherein the isolation block is used to record a block number of an isolated data block; and the data area comprises a data block used for file storage;
determining a to-be-accessed data block according to the request;
accessing a physical area corresponding to the data block on the disk;
acquiring a recorded average time consumed for accessing the disk;
calculating a ratio of time consumed for accessing the physical area to the average time consumed for accessing the disk;
comparing the ratio with a time consumption threshold preset in the file system, wherein if the ratio is greater than the time consumption threshold, the data block is located in the low-speed area on the disk, or if the ratio is less than the time consumption threshold, the data block is not located in the low-speed area on the disk;
recording, by the file system in the isolation block, a block number of the data block that is identified to be located in the low-speed area;
after subsequent access to the data block is received, determining whether the isolation block records the block number of the data block; and
after determining that the isolation block already records the block number of the data block, denying the subsequent access to the data block.

2. The method according to claim 1 , further comprising:
when the data block is not located in the low-speed area on the disk, updating, by the file system, the recorded average time consumed for accessing the disk.

3. The method according to claim 2 , wherein the file system further records a quantity of times the disk is accessed;
wherein the updating the recorded average time consumed for accessing the disk comprises:
calculating, by using a formula acctme(new)=(acctime(sb)*ion(sb)+acctime(file block))/(ion(sb)+1), an average time consumed for accessing the disk, wherein acctme(new) is the latest average time consumed for accessing the disk that is obtained by means of calculation, acctime(sb) is the average time consumed for accessing the disk that is recorded in the file system, ion(sb) is the quantity of times the disk is accessed that is recorded in the file system, and acctime(file block) is the time consumed for currently accessing the physical area; and
updating, by the file system, the recorded average time consumed for accessing the disk to the average time consumed for accessing the disk that is obtained by means of calculation; and
wherein the method further comprises: each time the average time consumed for accessing the disk is updated, increasing, by the file system, the quantity of times the disk is accessed by one.

4. The method according to claim 1 , wherein the data structure further comprises a metadata area, wherein:
the metadata area comprises an index node that is used for recording metadata of a file, wherein each file is corresponding to one index node; and
the isolating the data block that is identified to be located in the low-speed area further comprises:
setting a slow disk tag in an index node of the file, wherein the slow disk tag is used to indicate that the file occupies the data block that is located in the low-speed area on the disk; and
before the determining whether the isolation block records the block number of the data block, the method further comprises:
determining that the slow disk tag is set in the index node of the file.

5. The method according to claim 1 , wherein the data structure further comprises a superblock, wherein:
the superblock records the average time consumed for accessing the disk and the quantity of times the disk is accessed, and the superblock further records start addresses of the isolation block, the data area, and the metadata area, and a size of a data block.

6. The method according to claim 1 , wherein when the access is file reading, the determining the to-be-accessed data block comprises: determining, by the file system, the to-be-accessed data block based on a location of to-be-read data in the file and the size of the data block.

7. The method according to claim 1 , wherein when the access is file writing, the determining the to-be-accessed data block comprises:
determining, by the file system according to the request, whether a new data block needs to be allocated; and
if a new data block needs to be allocated, allocating, by the file system, a new data block to the file with skipping the isolated data block, and determining the new data block as the to-be-accessed data block; or
if a new data block does not need to be allocated, determining, by the file system, the to-be-accessed data block based on a write location in the file and the size of the data block.

8. A disk area isolation device, comprising:
a memory storing instructions of a file system; and
a processor coupled to the memory to execute the instructions of the file system to:
receive a request for access to a file, wherein the access comprises file reading or file writing,
wherein the file system comprises a data structure that is used to describe logical partitioning of the disk, and the data structure comprises an isolation block and a data area,
wherein the isolation block is used to record a block number of an isolated data block and the data area comprises a data block used for file storage:
determine a to-be-accessed data block according to the request;
access a physical area corresponding to the data block on the disk; acquire a recorded average time consumed for accessing the disk; calculate a ratio of time consumed for accessing the physical area to the average time consumed for accessing the disk;
compare the ratio with a time consumption threshold preset in the file system, wherein if the ratio is greater than the time consumption threshold, the data block is located in the low-speed area on the disk, or if the ratio is less than the time consumption threshold, the data block is not located in the low-speed area on the disk;
record, in the isolation block, a block number of the data block that is identified to be located in the low-speed area;
after subsequent access to the data block is received, determine whether the isolation block records the block number of the data block; and
after determining that the isolation block already records the block number of the data block, denying the subsequent access to the data block.

9. The device according to claim 8 , wherein the processor is further configured to: when the data block is not located in the low-speed area on the disk, update the recorded average time consumed for accessing the disk.

10. The device according to claim 9 , wherein the processor is further configured to record a quantity of times the disk is accessed;
wherein the processor updates the recorded average time consumed for accessing the disk comprises:
calculating, by the processing module by using a formula acctme(new)=(acctime(sb)*ion(sb)+acctime(file block))/(ion(sb)+1), an average time consumed for accessing the disk, wherein acctme(new) is the latest average time consumed for accessing the disk that is obtained by means of calculation, acctime(sb) is the average time consumed for accessing the disk that is recorded in the file system, ion(sb) is the qu

Assignees

• Huawei Tech Co Ltd

Inventors

• Shen Jiangang
• Li Yaohui

Classifications

• G06F3/06

  Digital input from, or digital output to, record carriers {, e.g. RAID, emulated record carriers or networked record carriers} · CPC title

• G06F3/0619

  in relation to data integrity, e.g. data losses, bit errors · CPC title

• G06F3/061 Primary

  Improving I/O performance · CPC title

• G06F3/0665

  at area level, e.g. provisioning of virtual or logical volumes · CPC title

• G06F3/0689

  Disk arrays, e.g. RAID, JBOD · CPC title