Nothing Special   »   [go: up one dir, main page]

CN1249581C - A hot backup data migration method - Google Patents

A hot backup data migration method Download PDF

Info

Publication number
CN1249581C
CN1249581C CN 02148818 CN02148818A CN1249581C CN 1249581 C CN1249581 C CN 1249581C CN 02148818 CN02148818 CN 02148818 CN 02148818 A CN02148818 A CN 02148818A CN 1249581 C CN1249581 C CN 1249581C
Authority
CN
China
Prior art keywords
data
hotspare disk
buffer memory
write
migtation
Prior art date
Legal status (The legal status 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 status listed.)
Expired - Fee Related
Application number
CN 02148818
Other languages
Chinese (zh)
Other versions
CN1501364A (en
Inventor
张巍
张国彬
任雷鸣
陈绍元
郑珉
胡鹏
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Huawei Technologies Co Ltd
Original Assignee
Huawei Technologies Co Ltd
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 Huawei Technologies Co Ltd filed Critical Huawei Technologies Co Ltd
Priority to CN 02148818 priority Critical patent/CN1249581C/en
Publication of CN1501364A publication Critical patent/CN1501364A/en
Application granted granted Critical
Publication of CN1249581C publication Critical patent/CN1249581C/en
Anticipated expiration legal-status Critical
Expired - Fee Related legal-status Critical Current

Links

Images

Landscapes

  • Memory System Of A Hierarchy Structure (AREA)

Abstract

The present invention relates to a hot backup disk data migration method which belongs to the field of disk storage. The method comprises the following steps: a. setting the length upper limit A of a disk I/O scheduling queue and triggering data migration; b. judging whether the length of the disk I/O scheduling queue is less than A; if yes, executing hot backup disk data migration; if no, stopping the data migration; c. executing mirror image copying for a data disk and a hot backup disk in the part completed the data migration until the end of the hot backup disk data migration. The present invention maintains the fixed position of the hot backup disk and simultaneously realizes the unified management of RAID group disks.

Description

A kind of HotSpare disk data migration method
Technical field
The present invention relates to the disk storage field, relate in particular to a kind of in disc array system (RAID:Redundant Array of Inexpensive Disks), the HotSpare disk data migration method.
Technical background
Along with the develop rapidly of science and technology and the widespread usage of computer technology, people are more and more higher to the performance requirement of memory device, and the RAID technology is widely used in the disk array as a proven technique.It is to combine with data strip blocking method by disk array, to improve a kind of structure of data available rate, realize directly reading of data from mirror-image copies of data redundancy by data image, system can automatically exchange on the mirrored disk, and does not need the data of recombinating and losing efficacy.RAID can be divided into RAID rank 1 to RAID rank 6, and be commonly referred to: RAID 0, and RAID 1, and RAID 2, and RAID 3, and RAID 4, and RAID 5.Each RAID rank all has the strengths and weaknesses of oneself." parity checking " is defined as the redundant information of user data, when hard disk lost efficacy, can produce data again.
To a RAID system with redundant ability, its disk array comprises data disk and HotSpare disk two parts.After certain data disk in the disk array lost efficacy, the RAID system will start its reconfiguration program, with the data reconstruction in the failed disk to HotSpare disk.Because being generally all RAID group membership disks (data disk), shares HotSpare disk, if through after the reconstruct several times, do not carry out data migtation from HotSpare disk to the RAID group membership disk (data disk) that upgrades, promptly after the fail data disk is replaced, not with in the data disk of the data migtation on the HotSpare disk after change, then RAID is organized all disks, the management that comprises member's disk and HotSpare disk brings great inconvenience (the HotSpare disk position is unfixing, but is scattered in the middle of member's disk).In addition, the method for some data migtation at the method CopyBack that RAIDFrame adopts, must be carried out data migtation by off-line such as the Carnegie Mellon University, that is to say, forbids having the user to ask to issue in data migration process.This does not obviously conform to online services in 24 * 7 hours of RAID system requirements.
At present, because the HotSpare disk data migration method is complicated, also do not see corresponding implementation method.
Summary of the invention
Purpose of the present invention just provides a kind of method of HotSpare disk data migtation, to solve the problem of spare drive data migtation.
A kind of HotSpare disk data migration method is characterized in that may further comprise the steps:
A, application cache pool are also set a HotSpare disk and are used to handle the magnetic disc i/o scheduling queue length upper limit A that comes from host requests, the trigger data migration;
B, judge HotSpare disk be used to handle come from host requests magnetic disc i/o scheduling queue length whether less than A, if, carry out the HotSpare disk data migtation, if not, then stop data migtation;
C, to finishing the part of data migtation, data disks and HotSpare disk are done mirror-image copies, finish until the HotSpare disk data migtation.
Described cache pool comprises free buffer memory and reserves buffer memory.
In the process of above-mentioned data migtation, the I/O scheduling queue priority that HotSpare disk and corresponding data dish thereof are used for data migtation is used to handle the I/O scheduling queue priority that comes from the main frame read and write access less than HotSpare disk and corresponding data dish thereof, also be, HotSpare disk and corresponding data dish thereof all answer priority processing to come from the read-write requests of main frame, are the request of data migtation and then.
In the process of above-mentioned data migtation, if the read request of main frame is arranged, then hang up the read request of main frame, data until this stripe unit of HotSpare disk are write in the free buffer memory fully, recover the main frame read request be suspended again, direct reading of data from free buffer memory is with the read request of response main frame to this stripe unit.
In the process of above-mentioned data migtation, when response main frame read request, the data migtation I/O scheduling queue of HotSpare disk is suspended.
In the process of above-mentioned data migtation, if the data of certain stripe unit of HotSpare disk write in the free buffer memory fully, waiting for that data migtation arrives data disks, if run into the read request of main frame this moment to this stripe unit, then hang up the data migtation of HotSpare disk, direct reading of data from free buffer memory, reading of data is recovered the data migtation of HotSpare disk afterwards again.
In the process of above-mentioned data migtation, if certain stripe unit of HotSpare disk is carrying out data migtation, be about to be read into data in buffer and be written to corresponding data disks, if run into the read request of main frame this moment to this stripe unit, direct reading of data from free buffer memory then is with the read request of response main frame to this stripe unit.
In the process of above-mentioned data migtation, if write to certain free buffer memory from the HotSpare disk sense data, if this moment, main frame proposed write request to this stripe unit, if also have the buffer memory of reserving, then the data that directly need write are write in the buffer memory of reservation; Exhaust if reserve buffer memory, then earlier to system's application one block cache, the data that need write are write in this buffer memory, the data migtation buffer memory of HotSpare disk is redirected to the buffer memory of new appointment.
In the process of above-mentioned data migtation, if the data of certain stripe unit of HotSpare disk write in the free buffer memory fully, waiting for that data migtation arrives data disks, if run into the write request of main frame this moment to this stripe unit, then hang up the data migtation of HotSpare disk, the required data that write of main frame are directly write in this free buffer memory, recovered the data migration process of HotSpare disk after finishing again.
In the process of above-mentioned data migtation, if certain stripe unit of HotSpare disk will be read into data in buffer and be written to corresponding data disks, if run into the write request of main frame to this stripe unit this moment, if also have the buffer memory of reserving, then the data that directly need write are write in the buffer memory of reservation; Exhaust if reserve buffer memory, then earlier to system's application one block cache, the data that need are write are write in this buffer memory again, after the data in the free buffer memory are finished by migration, again the data in the buffer memory of new appointment are carried out data migtation.
The present invention has used for reference towards the architecture of disk restructing algorithm, virtual non-volatile Cache and design proposal, when keeping the HotSpare disk stationkeeping, realizes the unified management of RAID group disk.
Description of drawings
Fig. 1 is a data migtation process flow diagram of the present invention;
Fig. 2 is in the data migration process of the present invention, the synoptic diagram of mirror-image copies;
Fig. 3 is that data migtation of the present invention and write request concern synoptic diagram;
Fig. 4 is that data migtation of the present invention and read request concern synoptic diagram;
Fig. 5 is the processing synoptic diagram of a data migration of the present invention critical conditions main frame read request;
Fig. 6 is the processing synoptic diagram of a data migration of the present invention critical conditions main frame write request.
Embodiment
Below in conjunction with Figure of description the specific embodiment of the present invention is described.
So-called HotSpare disk data migtation, promptly after the disk that lost efficacy is replaced by new building, the HotSpare disk general is automatically on the data disks of data migtation to the original position.This HotSpare disk data migtation is that internal system realizes, removing needs manual activation, transition process to need not to interfere when starting.
Adopt HotSpare disk data migration method of the present invention, as shown in Figure 1, may further comprise the steps:
Magnetic disc i/o scheduling queue length upper limit A that a, HotSpare disk of setting are used to handle come from host requests and trigger data migration;
The data migration process of this disk should be carried out in the free time of HotSpare disk.The so-called free time is meant that promptly the I/O scheduling queue length QueueLength of HotSpare disk is 0 if (for example 50ms) do not have the access request of disk within a certain period of time, and disk just is considered to be in idle phase, and data migration process just can begin.
This process need triggers artificially, after failed disk is replaced, and operator's manual triggers data migtation.After data migtation starts, but the manual cancellation.In whole transition process, the progress of record data migration constantly so that support breakpoint transmission, promptly after system cut-off is restarted, can continue migration from breakpoint.This whole transition process is handled by CPU, and is controlled by primary control program.
In fact, be difficult to accomplish that the I/O scheduling queue length QueueLength of HotSpare disk is 0, therefore, we are provided with the upper limit A of an I/O scheduling queue length QueueLength, as for the value principle of A, relevant with the I/O scheduling queue maximum length Max_QueueLength of HotSpare disk.Generally speaking, A can be decided to be 10% of Max_QueueLength, or littler, can as long as satisfy system requirements.
Continuous for the data migtation that makes HotSpare disk as far as possible, data migtation can be arranged in visit capacity and carry out little the time, such as the late into the night.
During the HotSpare disk data migtation because the speed of write data dish must be less than reading HotSpare disk speed, so before the data migtation pre-first to file one block cache pond guarantee to read HotSpare disk continue carry out.Data migration process can be regarded as the process of two independent parallels: read HotSpare disk process and write data dish process, read the HotSpare disk process and the data of HotSpare disk are read in the cache pool temporary by stripe unit, write data dish process will read the data cached data disks that writes of end.
In the present invention, cache pool comprises free buffer memory and reserves buffer memory two parts that free cache size is 4~5 stripe unit sizes, is used for normal data migtation, reserving cache size is 2~3 stripe unit sizes, is used for the main frame write request data of temporal data transition process.Because in the RAID group, the stripe unit size may be inconsistent, for ease of unified internal memory application, the size of the every block cache in the cache pool is determined by the segment depth of stripe unit maximum.
B, judge HotSpare disk be used to handle come from host requests magnetic disc i/o scheduling queue length whether less than A, if, carry out the HotSpare disk data migtation, if otherwise stop data migtation;
When magnetic disc i/o scheduling queue length QueueLength<A, begin to carry out the data migtation of HotSpare disk, otherwise stop data migtation.
C, to finishing the part of data migtation, data disks and HotSpare disk are done mirror-image copies, finish until the HotSpare disk data migtation.
As shown in Figure 2, carrying out along with the HotSpare disk data migtation, to finishing the write access of data migtation (being in the corresponding with it stripe unit of data written data dish of HotSpare disk) part, must do mirror-image copies to data disks and HotSpare disk, be about to the write request that main frame is sent, be sent to HotSpare disk and data disks simultaneously.Make like this that in data migtation data disks can keep the consistent of data all the time with HotSpare disk, and the up-to-date data of the complete all the time maintenance of the data on the HotSpare disk, when losing efficacy with data disks in the convenient data migration process, the data on the HotSpare disk are still available.
As shown in Figure 3, if the main frame write request only is sent to data disks, then data available will be distributed on HotSpare disk and the data disks, in case data disks lost efficacy, will cause the part latest data to need reconstruct to recover.
As shown in Figure 4, to the part of not doing data migtation in the HotSpare disk as yet and to finishing the read access of data migration portion, then directly the corresponding stripe unit of HotSpare disk is carried out read-write operation.
In above-mentioned data migration process, may run into following several critical condition, the present invention takes corresponding disposal route:
1, main frame is a read request:
(a) when writing fashionable to certain free buffer memory from the HotSpare disk sense data, if run into the read request of main frame to this stripe unit, but the read request of temporary suspension main frame, data until this stripe unit of HotSpare disk are write in the free buffer memory fully, recover the main frame read request be suspended again, direct reading of data from free buffer memory is with the read request of response main frame to this stripe unit.In response main frame read request process, the data migtation I/O scheduling queue of HotSpare disk is suspended.
(b) data of certain stripe unit of HotSpare disk write in the free buffer memory fully, waiting for that data migtation arrives data disks, at this moment, if run into the read request of main frame to this stripe unit, can hang up the data migtation of HotSpare disk, direct reading of data from free buffer memory, with the read request of response main frame to this stripe unit, and then the data migration process of recovery HotSpare disk.
As shown in Figure 5, promptly be the synoptic diagram of this kind situation, among the figure
(1): sense data is in free buffer memory from HotSpare disk;
(2): the read request of main frame;
(3): the read request of main frame is redirected to free buffer memory.
(c) certain stripe unit of HotSpare disk will be read into data in buffer and be written to corresponding data disks, at this moment, if run into the read request of main frame to this stripe unit, can be from free buffer memory direct reading of data, with the read request of response main frame to this stripe unit.
2, main frame is a write request:
(a) write to certain free buffer memory from the HotSpare disk sense data.At this moment, if run into the write request of main frame to this stripe unit, if also have the buffer memory of reserving, then the data that directly need write are write in the buffer memory of reservation; Exhaust if reserve buffer memory, then earlier to system's application one block cache, the data that need are write are write in this buffer memory again.And the data in the used free buffer memory of this stripe unit are cancelled.The data migtation buffer memory of HotSpare disk is redirected to the buffer memory of new appointment.
(b) data of certain stripe unit of HotSpare disk write in the free buffer memory fully, waiting for that data migtation arrives data disks, at this moment, if run into the write request of main frame to this stripe unit, can hang up the data migtation of HotSpare disk, the required data that write of main frame are directly write in this free buffer memory.And then the data migration process of recovery HotSpare disk.
(c) certain stripe unit of HotSpare disk will be read into data in buffer and be written to corresponding data disks, and at this moment, if run into the write request of main frame to this stripe unit, if also have the buffer memory of reserving, then the data that directly need write are write in the buffer memory of reservation; Exhaust if reserve buffer memory, then earlier to system's application one block cache, the data that need are write are write in this buffer memory again.After the data in the free buffer memory are finished by migration, again the data in the buffer memory of new appointment are carried out data migtation.Because on the RAID group M LUN (Logical Unit Number) arranged, and N itemize arranged on each LUN, corresponding certain disk in this RAID group then has M * N stripe unit.During data migtation, just, move successively from small to large, finish until all stripe unit migrations according to the stripe unit address on this disk.So-called stripe unit is carrying out data migtation, is meant that the data of this stripe unit are being read buffer memory, or has been read buffer memory, waits for being written into corresponding data disks, or the data in the buffer memory is write corresponding data disks.
As shown in Figure 6, promptly be the synoptic diagram of this kind situation, among the figure:
(1) from HotSpare disk sense data in free buffer memory;
(2) will be read into free data in buffer and be written to corresponding data disks;
(3) write request of main frame;
(4) write request of main frame is redirected to the buffer memory of reserving buffer memory or applying for again;
(5) wait to finish free data in buffer and write after, carry out data migtation with the host data in the buffer memory of reserving buffer memory or application again again.
When the data migtation of write operation, must carry out mirror-image copies to the stripe unit of HotSpare disk and data disks correspondence, also promptly must simultaneously corresponding data be write in the data disks of HotSpare disk and correspondence, to guarantee the consistance of HotSpare disk and data of data discs.
The above ground HotSpare disk data migtation buffer memory quantity can be set to 4~5, and reserves 2~3 and come out standby.Because the speed of write data dish must be less than reading HotSpare disk speed, and consider the visit of external host to data disks, writing the dish time will be much higher than and read to coil the used time.So under the prerequisite that system allows, can suitably increase HotSpare disk data migtation buffer memory quantity.In order to deal with the critical condition of data migtation, reserve 2~3 buffer memorys and use to being in critical data migration stripe unit specially in addition.
When the HotSpare disk data migtation, but first to file one block cache pond, and the cache pool size is 4~5 stripe unit sizes.Because in the RAID group, the stripe unit size may be inconsistent, for ease of unified internal memory application, the size of the every block cache in the cache pool is determined by the segment depth of stripe unit maximum.
The present invention has used for reference towards the architecture of disk restructing algorithm, virtual non-volatile Cache and design proposal; Keep the HotSpare disk stationkeeping, realized the unified management of RAID group disk.
The above; only for the preferable embodiment of the present invention, but protection scope of the present invention is not limited thereto, and anyly is familiar with those skilled in the art in the technical scope that the present invention discloses; the variation that can expect easily or replacement all should be encompassed within protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with the protection domain of claims.

Claims (10)

1, a kind of HotSpare disk data migration method is characterized in that may further comprise the steps:
A, application cache pool are also set a HotSpare disk and are used to handle the magnetic disc i/o scheduling queue length upper limit A that comes from host requests, the trigger data migration;
B, judge HotSpare disk be used to handle come from host requests magnetic disc i/o scheduling queue length whether less than A, if, carry out the HotSpare disk data migtation, if otherwise stop data migtation;
C, to finishing the part of data migtation, data disks and HotSpare disk are done mirror-image copies, finish until the HotSpare disk data migtation.
2, HotSpare disk data migration method as claimed in claim 1 is characterized in that comprising free buffer memory and reserving buffer memory at described cache pool.
3, HotSpare disk data migration method as claimed in claim 1 or 2, it is characterized in that in the process of above-mentioned data migtation the magnetic disc i/o scheduling queue priority that HotSpare disk and corresponding data dish thereof are used for data migtation is used to handle the magnetic disc i/o scheduling queue priority that comes from the main frame read and write access less than HotSpare disk and corresponding data dish thereof.
4, HotSpare disk data migration method as claimed in claim 3, it is characterized in that in the process of above-mentioned data migtation, if write to certain free buffer memory from the HotSpare disk sense data, if the read request of main frame is arranged, then hang up the read request of main frame, write fully in the free buffer memory, recover the main frame read request that is suspended again until the data of this stripe unit of HotSpare disk, direct reading of data from free buffer memory is with the read request of response main frame to this stripe unit.
5, HotSpare disk data migration method as claimed in claim 4 is characterized in that in the process of above-mentioned data migtation, and when response main frame read request, the data migtation I/O scheduling queue of HotSpare disk is suspended.
6, HotSpare disk data migration method as claimed in claim 3, it is characterized in that in the process of above-mentioned data migtation, if the data of certain stripe unit of HotSpare disk write in the free buffer memory fully, waiting for that data migtation arrives data disks, if run into the read request of main frame this moment to this stripe unit, then hang up the data migtation of HotSpare disk, direct reading of data from free buffer memory, reading of data is recovered the data migtation of HotSpare disk afterwards again.
7, HotSpare disk data migration method as claimed in claim 3, it is characterized in that in the process of above-mentioned data migtation, if certain stripe unit of HotSpare disk will be read into data in buffer and be written to corresponding data disks, if run into the read request of main frame this moment to this stripe unit, direct reading of data from free buffer memory then is with the read request of response main frame to this stripe unit.
8, HotSpare disk data migration method as claimed in claim 3, it is characterized in that in the process of above-mentioned data migtation, if write to certain free buffer memory from the HotSpare disk sense data, if this moment, main frame was to this stripe unit write request, if also have the buffer memory of reserving, then the data that directly need write are write in the buffer memory of reservation; Exhaust if reserve buffer memory, then earlier to system's application one block cache, the data that need write are write in this buffer memory, the data migtation buffer memory of HotSpare disk is redirected to the buffer memory of new appointment.
9, HotSpare disk data migration method as claimed in claim 3, it is characterized in that in the process of above-mentioned data migtation, if the data of certain stripe unit of HotSpare disk write in the free buffer memory fully, waiting for that data migtation arrives data disks, if run into the write request of main frame this moment to this stripe unit, then hang up the data migtation of HotSpare disk, the required data that write of main frame are directly write in this free buffer memory, recover the data migration process of HotSpare disk after finishing again.
10, HotSpare disk data migration method as claimed in claim 3, it is characterized in that in the process of above-mentioned data migtation, if certain stripe unit of HotSpare disk will be read into data in buffer and be written to corresponding data disks, if run into the write request of main frame this moment to this stripe unit, if also have the buffer memory of reserving, then the data that directly need write are write in the buffer memory of reservation; Exhaust if reserve buffer memory, then earlier to system's application one block cache, the data that need are write are write in the buffer memory of reservation again, after the data in the free buffer memory are finished by migration, again the data in the buffer memory of new appointment are carried out data migtation.
CN 02148818 2002-11-18 2002-11-18 A hot backup data migration method Expired - Fee Related CN1249581C (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN 02148818 CN1249581C (en) 2002-11-18 2002-11-18 A hot backup data migration method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN 02148818 CN1249581C (en) 2002-11-18 2002-11-18 A hot backup data migration method

Publications (2)

Publication Number Publication Date
CN1501364A CN1501364A (en) 2004-06-02
CN1249581C true CN1249581C (en) 2006-04-05

Family

ID=34233347

Family Applications (1)

Application Number Title Priority Date Filing Date
CN 02148818 Expired - Fee Related CN1249581C (en) 2002-11-18 2002-11-18 A hot backup data migration method

Country Status (1)

Country Link
CN (1) CN1249581C (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107844590A (en) * 2017-11-17 2018-03-27 郑州云海信息技术有限公司 Small documents moving method, device, equipment and computer-readable recording medium

Families Citing this family (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4387261B2 (en) * 2004-07-15 2009-12-16 株式会社日立製作所 Computer system and storage system migration method
EP2021903A2 (en) * 2006-05-24 2009-02-11 Compellent Technologies Data progression disk locality optimization system and method
CN101207513B (en) * 2006-12-22 2012-09-05 中兴通讯股份有限公司 Apparatus and method for saving historical data
CN100451942C (en) * 2007-02-05 2009-01-14 威盛电子股份有限公司 Data migration method between the independent storage device extended and adjusted storage device
TWI334564B (en) 2007-02-14 2010-12-11 Via Tech Inc Data migration systems and methods for independent storage device expansion and adaptation
CN101814010B (en) * 2010-01-11 2011-11-23 浪潮(北京)电子信息产业有限公司 Disk array implementation method and device
CN102521058A (en) * 2011-12-01 2012-06-27 北京威视数据系统有限公司 Disk data pre-migration method of RAID (Redundant Array of Independent Disks) group
CN102591746B (en) * 2011-12-29 2014-01-08 华为数字技术(成都)有限公司 Data reconstruction method and storage equipment
CN102880428B (en) * 2012-08-20 2015-09-09 华为技术有限公司 The creation method of distributed Redundant Array of Independent Disks (RAID) and device
CN104615511B (en) * 2015-02-05 2018-01-26 中国工商银行股份有限公司 One kind recovers processing method and processing device based on bicentric host batch
CN109213422B (en) * 2017-06-30 2021-10-01 南宁富桂精密工业有限公司 Hierarchical storage method, hierarchical storage device, and computer-readable storage medium
CN112698794A (en) * 2021-02-02 2021-04-23 南京群顶科技有限公司 Method and device for realizing online migration of distributed storage volume

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107844590A (en) * 2017-11-17 2018-03-27 郑州云海信息技术有限公司 Small documents moving method, device, equipment and computer-readable recording medium
CN107844590B (en) * 2017-11-17 2020-05-29 郑州云海信息技术有限公司 Small file migration method, device, equipment and computer readable storage medium

Also Published As

Publication number Publication date
CN1501364A (en) 2004-06-02

Similar Documents

Publication Publication Date Title
US7975168B2 (en) Storage system executing parallel correction write
US7774540B2 (en) Storage system and method for opportunistic write-verify
US9665442B2 (en) Smart flushing of data to backup storage
US7958310B2 (en) Apparatus, system, and method for selecting a space efficient repository
CN1249581C (en) A hot backup data migration method
CN104350477B (en) For solid-state drive device(SSD)Optimization context remove
US7137038B2 (en) System and method for autonomous data scrubbing in a hard disk drive
US7047358B2 (en) High-performance log-structured RAID
US8356126B2 (en) Command-coalescing RAID controller
US5537588A (en) Partitioned log-structured file system and methods for operating the same
KR101824295B1 (en) Cache management including solid state device virtualization
US8443134B2 (en) Apparatus, system, and method for graceful cache device degradation
JP3697149B2 (en) How to manage cache memory
JP2001296974A (en) System and method for storing data
US5600816A (en) System and method for managing data in a cache system for a disk array
JP2008276646A (en) Storage device and data management method for storage device
US20040205297A1 (en) Method of cache collision avoidance in the presence of a periodic cache aging algorithm
JP2002323959A (en) System and method for non-volatile write cache based on log of magnetic disk controller
US20090265578A1 (en) Full Stripe Processing for a Redundant Array of Disk Drives
CN1848071A (en) Computer system, disk apparatus and data update control method
US6721863B1 (en) Disk control mechanism preferable for random disk write
US20130024618A1 (en) Log structure array
US9785438B1 (en) Media cache cleaning based on workload
US20120102242A1 (en) Controlling data destaging within a multi-tiered storage system
US20190243578A1 (en) Memory buffer management for solid state drives

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20060405

Termination date: 20181118

CF01 Termination of patent right due to non-payment of annual fee