WO2010075703A1 - 分布式文件系统中多磁盘分组热备的方法与装置 - Google Patents

分布式文件系统中多磁盘分组热备的方法与装置 Download PDF

Info

Publication number
WO2010075703A1
WO2010075703A1 PCT/CN2009/073866 CN2009073866W WO2010075703A1 WO 2010075703 A1 WO2010075703 A1 WO 2010075703A1 CN 2009073866 W CN2009073866 W CN 2009073866W WO 2010075703 A1 WO2010075703 A1 WO 2010075703A1
Authority
WO
WIPO (PCT)
Prior art keywords
disk
hot standby
group
unit
file
Prior art date
Application number
PCT/CN2009/073866
Other languages
English (en)
French (fr)
Inventor
朱红军
刘洋
Original Assignee
中兴通讯股份有限公司
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 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Priority to EP09835990.4A priority Critical patent/EP2378434B1/en
Publication of WO2010075703A1 publication Critical patent/WO2010075703A1/zh

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00Error detection; Error correction; Monitoring
    • G06F11/07Responding to the occurrence of a fault, e.g. fault tolerance
    • G06F11/16Error detection or correction of the data by redundancy in hardware
    • G06F11/20Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements
    • G06F11/2053Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements where persistent mass storage functionality or persistent mass storage control functionality is redundant
    • G06F11/2094Redundant storage or storage space

Definitions

  • the present invention relates to a method and apparatus for multi-disk packet hot standby in a distributed file system (DFS).
  • DFS distributed file system
  • a file access server In DFS, a file access server (File Access Sever, called FAS) is provided. Each FAS provides file file read and write functions, and each FAS manages multiple disks. The FAS is timed to the file locating register (File Locating Register), and the disk discovery function is also implemented by the FAS.
  • a disk backup method is generally adopted to improve the reliability of the system and reduce the impact of FAS failure on the entire system, that is, when a single disk fails, a backup disk is used to continue to provide services.
  • the main object of the present invention is to provide a method and apparatus for multi-disk group hot standby in a distributed file system, which can effectively improve the reliability of the DFS.
  • a method for hot-standby multi-disk grouping in a distributed file system includes:
  • the file access server allocates disks to the object, and establishes a hot standby relationship between the disks by using the group as a target.
  • the disk is powered on, it is determined whether the powered disk is a new disk, and if so, according to the new disk.
  • the group determines a hot standby relationship of the new disk, and loads a backup file for the new disk; If it is not a new disk, it is determined whether the current group of the power-on disk is the same as the original group.
  • the original hot-standby relationship of the power-on disk is restored, and the file content of the power-on disk is determined. Whether the contents of the files on the disk are the same. If they are the same, they are not processed. If the backup file is not loaded for the power-on disk, if the data is inconsistent, the hot disk is re-established according to the current group of the power-on disk. Prepare the relationship and reload the backup file for the powered disk.
  • the grouping of the disks by the file access server is specifically: configuring physical information, logical unit information, and group number information of the location of the file access server, establishing a disk parameter and the The correspondence between the physical information of the file access location, the logical unit information, and the packet number information.
  • the physical information of the location includes information about a shelf, a port, and a slot location of the file access server;
  • the logical unit information includes a CPU, an internal port unit, and an evening of the file access server.
  • the disk parameter includes a disk data area usage identifier, a logical address index number of the disk in the file access server, a physical serial number of the disk, a logical identifier of the disk, and a check query of the disk block Serial number, disk heartbeat detection, disk status, total disk capacity, disk free capacity, amount of data read by the disk for a certain period of time, amount of data written by the disk for a certain period of time, start time of •t ⁇ , end time of 4 ⁇ .
  • the hot standby relationship between the disks is set up by the group, and the hot standby number is set for the disks in the group, and the dynamic correspondence between the disks between the groups is established according to the hot standby number.
  • an apparatus for multi-disk packet hot standby in a distributed file system is provided.
  • the device for multi-disk group hot standby in the distributed file system comprises: a disk grouping unit for grouping disks with a file access server; a hot standby relationship static configuration unit, for grouping Establishing a hot standby relationship between the disks; a monitoring unit for monitoring the updated status of the disk; a new disk detecting unit for determining whether the powered disk is a new disk, and if so, triggering a new disk hot standby relationship determining unit, if Otherwise, the group judgment unit is triggered; a new disk hot standby relationship determining unit, configured to determine a hot standby relationship of the new disk according to the group in which the new disk is located, and trigger a backup file loading unit; and a group determining unit, configured to determine the current powering disk Whether the group is the same as the original group.
  • the hot standby relationship recovery and file content judgment unit are triggered. If they are inconsistent, the hot standby relationship update unit hot standby relationship recovery and file content judgment unit is triggered to restore the powered disk.
  • the hot standby relationship is determined, and it is determined whether the file content of the powered disk is the same as the file content of the hot spare disk. If the same, the file is not processed. If different, the backup file loading unit is triggered; the hot standby relationship updating unit is used for The backup file loading unit is triggered by the backup file loading unit, and the backup file loading unit is configured to load the backup file for the power-on disk according to the current group of the power-on disk.
  • the disk grouping unit groups the disks by using the file access server as an object, specifically: configuring physical information, logical unit information, and group number information of the location of the file access server, and establishing a disk parameter and a location.
  • the physical information of the location includes information about a shelf, a port, and a slot location of the file access server; the logical unit information includes a CPU, an internal port unit, and an evening of the file access server.
  • the disk parameter includes a disk data area usage identifier, a logical address index number of the disk in the file access server, a physical serial number of the disk, a logical identifier of the disk, and a check query of the disk block Serial number, disk heartbeat detection, disk status, total disk capacity, disk free capacity, amount of data read by the disk for a certain period of time, amount of data written by the disk for a certain period of time, start time of •t ⁇ , end time of 4 ⁇ .
  • the hot standby relationship static configuration unit establishes a hot standby relationship between the disks by using the group as a target, specifically: setting a hot spare number for the disks in the group, and establishing a disk between the groups according to the hot standby number. Dynamic correspondence. According to still another aspect of the present invention, a multi-disk grouping heat in a distributed file system is provided. The method of preparation.
  • the method for the hot standby of the multi-disk grouping in the distributed file system includes: grouping the disks with the file access server as an object, and establishing a hot standby relationship between the disks by using the group as a target; determining whether the powered disk is a new disk, If yes, the hot standby relationship of the new disk is determined according to the group in which the new disk is located, and the backup file is loaded for the new disk. Otherwise, according to whether the current group and the original group of the powered disk are Configure the hot standby relationship of the powered disk.
  • the hot standby relationship of the power-on disk is configured according to whether the current group of the power-on disk is in the same group as the original group: If the group is the same as the original group, restore the original hot standby relationship of the powered disk, and determine whether the file content of the powered disk is the same as the file content of the hot spare disk. If they are the same, they are not processed.
  • the backup file is reloaded for the power-on disk. If the current group of the power-on disk is inconsistent with the original group, the hot standby relationship is established for the power-on disk. And reload the backup file for the powered disk.
  • an apparatus for multi-disk packet hot standby in a distributed file system includes: a pre-configuration unit, configured to group disks by using a file access server, and establish a hot standby relationship between disks by using the group as an object.
  • the new disk detection unit is configured to determine whether the power-on disk is a new disk, and if so, trigger a new disk hot standby relationship determining unit; otherwise, trigger the old disk hot standby relationship configuration unit; Determining the hot standby relationship of the new disk according to the group in which the new disk is located, and loading a backup file for the new disk; the old disk hot standby relationship configuration unit is configured to use the powered disk Configure the hot standby relationship of the powered-on disk in the same group as the original one.
  • the old disk hot standby relationship configuration unit includes: a group determining subunit, configured to determine whether the current group of the power-on disk is consistent with the original group, and if yes, triggering the hot standby relationship recovery sub-unit, if the inconsistency
  • the hot standby relationship update subunit is triggered, and the hot standby relationship recovery subunit is configured to restore the original hot standby relationship of the powered disk, and determine the file content of the powered disk and the file content of the hot spare disk. If they are the same, if they are the same, they are not processed. If they are different, the backup file is reloaded for the powered disk.
  • the hot standby relationship update subunit is used to re-send the current group of the powered disk.
  • the hot disk establishes a hot standby relationship and reloads the backup file for the powered disk.
  • the invention uses FAS as the basis of disk grouping, all disks in one FAS are grouped into groups, and hot standby is performed in units of groups, so that whether there is a problem with a single disk or a single FAS problem, the entire DFS is not caused. influences.
  • the group is magnetic
  • the hot standby relationship between disks is not fixed, but dynamically changes according to the update status of the disk.
  • the present invention can automatically generate new ones according to the hot standby relationship between the original groups.
  • FIG. 1B is a schematic diagram of a method for hot standby of multiple disk packets in a distributed file system according to an embodiment of the present invention
  • 2A is a block diagram of a device for a multi-disk grouping device in a distributed file system according to an embodiment of the present invention
  • FIG. 2B is a schematic diagram of an old disk in a device for a multi-disk grouping device in a distributed file system according to an embodiment of the present invention
  • the detailed structure of the hot standby relationship configuration unit is shown in FIG. 2C.
  • FIG. 2C is a detailed structural block diagram of an apparatus for multi-disk group hot standby in a distributed file system according to an embodiment of the present invention.
  • the basic idea of the present invention is: With FAS as the basis of disk grouping, all disks in one FAS are grouped into one group, and hot standby is performed in units of groups, so that whether there is a problem with a single disk or There is a problem with a single FAS that does not affect the entire DFS.
  • the hot standby relationship between the disks in the group is not fixed, but dynamically changes according to the update status of the disk. When the disk in the FAS is powered on, it can be used.
  • FIG. 1A is a schematic flowchart of a method for hot standby of multi-disk grouping in a distributed file system according to an embodiment of the present invention. As shown in FIG. 1A, the method includes: Step 101a: grouping disks by using a file access server as an object, Set the hot standby relationship between the disks as the group. Step 102a: Determine whether the powered disk is a new disk.
  • the hot standby relationship of the powered disk is configured according to whether the current group of the powered disk is the same as the original group.
  • the hot-standby relationship between the power-on disk and the original group is the same as the original group.
  • restore the original hot standby relationship of the powered disk and determine whether the file content of the powered disk is the same as the content of the file on the hot spare disk. If they are the same, they are not processed. If different, the backup file is loaded for the powered disk.
  • FIG. 1B is a schematic flowchart of a method for hot standby of multi-disk grouping in a distributed file system according to an embodiment of the present invention.
  • the method includes: Step 101b: grouping disks by using a file access server Create a hot standby relationship between disks based on the group.
  • a data area for storing disk information is set on the FLR.
  • the disk information herein may include basic information of the disk, information about the FAS to which the disk belongs, and hot standby relationship between the disks in the group.
  • the data area of the FLR is physical information, logical unit information, and group number information of the FAS configuration location, and the correspondence between the disk parameters and the physical information of the FAS location, the logical unit information, and the packet number information is established.
  • the physical information of the location may include information about a shelf, a frame, a slot location of the file access server, and the like;
  • the logical unit information may include a CPU, an internal port unit, an external port unit, a MAC address, and an IP address of the FAS.
  • the disk parameters may include a disk data area usage identifier, a logical address index number of the disk in the file access server, a physical serial number of the disk, and a disk logical ID, disk block check query serial number, disk heartbeat detection, disk status, total disk capacity, disk free capacity, amount of data read by the disk for a certain period of time, amount of data written by the disk for a certain period of time, start time of 4 ⁇ , upper ⁇ ⁇ End time, etc.
  • all the disks in one FAS are grouped, and a hot standby relationship is established between the two groups, and a hot spare number is set for the disks in the group, and disks between the groups are established according to the hot standby number.
  • the hot standby correspondence between the disks in the group is dynamic: the number of hot spare numbers in each group is set according to the maximum number of disks supported by the FAS, in the FAS.
  • the hot standby method provided by the embodiment of the present invention can still be dynamically adjusted.
  • two groups that are hot standby are assigned the same group number, so as to be unique. That is to say, in the embodiment of the present invention, the hot standby relationship is limited to the group, and the hot standby relationship between the disks in the group can be arbitrarily set.
  • Step 102b When it is detected that the disk is powered on, determine whether the powered disk is a new disk.
  • the FAS periodically detects the disk update status in the FAS.
  • the detection result is 4 ⁇ to FLR, and the FLR is based on the FAS.
  • the result of the report is to modify the FLR data area, add or delete information about the disk, and so on.
  • the embodiment of the present invention is directed to a dynamic hot standby method of a disk. Therefore, the disk power-on is emphasized.
  • the disk is out of the FLR, the related configuration information of the disk is retained for a period of time, after the preset preset time is exceeded. , delete the relevant configuration information of the disk.
  • the FAS dynamically detects the parameters of the disk, including the physical serial number of the disk, the total disk capacity, and the disk free capacity. For details about the parameters, see the related description in step 101b. Again, the FAS reports the parameter information of the detected disk.
  • the FLR retrieves the data area according to the received disk parameter information, and checks whether the disk is a new disk. If yes, step 103b is performed; otherwise, step 104b is performed. Step 103b: Determine the hot standby relationship of the new disk according to the group where the new disk is located, and load the backup file for the new disk.
  • the FLR determines the group to which the new disk belongs according to the FAS that reports the disk parameter information, assigns a hot spare number to the new disk according to the group information, and determines another group that is hot standby with the group, as described above. Determine the two groups that are hot standby by the group number, and determine the disks in the other group that are not hot standby. If there are two, select one of the disks that are not hot standby to establish a mutual relationship with the new disk. For a hot standby relationship, if there is one, the disk that is not hot standby is established with the new disk. The hot standby relationship is not processed if it is not.
  • the hot standby relationship is associated with the hot spare number between the disk that is not hot standby and the new disk, and is stored in the data area of the FLR. After the hot standby relationship is established, the files in the original hot spare disk are copied to the new disk, and the process ends.
  • Step 104b Determine whether the current group of the power-on disk is consistent with the original group. If yes, go to step 105b. Otherwise, go to step 106b.
  • the FLR determines the group to which the power-on disk belongs originally according to the parameter information of the disk, determines whether the power-on disk has replaced the FAS, and if the FAS is replaced, the power-on disk is replaced from another FAS. Step 106b is performed. Otherwise, the powered disk is used for re-powering the disk in the FAS, and step 105b is performed. Step 105b: Recover the original hot standby relationship between the disks, and determine whether the file content of the powered disk is the same as the content of the file on the hot spare disk. If the file is the same, the processing is not performed. If not, the disk is loaded again. backup file.
  • the original disk is inserted in the FAS. It is very likely that the disk in the FAS is temporarily removed due to virtual interface connection, short-term maintenance, etc., and is inserted into the FAS.
  • FLR finds whether the hot standby relationship of the disk still exists. If the content of the file on the disk is consistent with the content of the file on the hot spare disk, the file is not updated. Otherwise, the file is updated to make the newly inserted disk and the file on the hot spare disk consistent. If the hot standby relationship of the disk is not available, restore the original hot standby relationship of the disk, and check whether the file content on the disk is consistent with the file content on the hot spare disk. ⁇ ) The content of the file in the newly inserted disk and the hot spare disk is updated to be consistent.
  • Step 106b Re-establish a hot standby relationship for the powered disk according to the current group of the powered disk, and reload the backup file for the powered disk.
  • the newly inserted disk is processed as a new disk, that is, the hot standby process is performed in the manner of step 103b, and details are not described herein again.
  • the content of the file in the newly inserted disk and the hot spare disk is updated.
  • the original hot standby configuration information of the powered disk is deleted. End the process.
  • FIG. 2A is a structural block diagram of an apparatus for a multi-disk grouping device in a distributed file system according to an embodiment of the present invention. As shown in FIG.
  • a device for multi-disk group hot standby in a distributed file system includes
  • the pre-configuration unit 12 is configured to group the disks by using the file access server, and establish a hot standby relationship between the disks by using the group as a target;
  • the new disk detecting unit 14 is connected to the pre-configuration unit 12, and is configured to Determining whether the power-on disk is a new disk, if yes, triggering the new disk hot standby relationship determining unit 16, otherwise, triggering the old disk hot standby relationship configuration unit 18;
  • the new disk hot standby relationship determining unit 16 is connected to the new disk detecting unit 14,
  • the hot spare relationship of the new disk is determined according to the group of the new disk, and the backup file is loaded for the new disk;
  • the old disk hot standby relationship configuration unit 18 is connected to the new disk detecting unit 14 for the power disk.
  • the hot standby relationship between the current disk and the original group is configured.
  • 2B is a block diagram showing a detailed structure of an old disk hot standby relationship configuration unit in a device of a multi-disk grouping device in a distributed file system according to an embodiment of the present invention.
  • the old disk hot standby relationship configuration unit 18 includes: The subunit 182 is configured to determine whether the current group of the power-on disk is consistent with the original group, and if yes, the hot standby relationship recovery sub-unit is triggered.
  • the hot standby relationship update sub-unit is triggered; the hot standby relationship recovery sub-unit 184,
  • the group determining unit 182 is configured to restore the original hot standby relationship of the powered disk, and determine whether the file content of the powered disk is the same as the content of the file on the hot spare disk. If the information is the same, the processing is not processed.
  • the power-on disk is loaded with the backup file; the hot standby relationship update sub-unit 186 is connected to the group judgment sub-unit 182, and is configured to re-establish a hot standby relationship for the powered-on disk according to the current group of the powered-on disk, and re-load the backup for the powered-on disk. file.
  • the apparatus for multi-disk group hot standby in a distributed file system includes a disk.
  • the grouping unit 20, the hot standby relationship static configuration unit 21, the monitoring unit 22, the new disk detecting unit 23, the new disk hot standby relationship determining unit 24, the group determining unit 25, the hot standby relationship recovery and file content determining unit 26, and the hot standby relationship The updating unit 27 and the backup file loading unit 28, wherein the disk grouping unit 20 is configured to group the disks with the file access server as objects, that is, all the disks in one FAS are grouped into one group.
  • the hot standby relationship static configuration unit 21 is configured to establish a hot standby relationship between the disks, that is, to establish a hot standby relationship between the groups, but does not limit the hot standby correspondence between the disks in the group.
  • the hot standby relationship group uses the same group number. For the establishment of the hot standby relationship, the basic parameter information of the disk and the FAS related information need to be reported.
  • the hot standby relationship is the relationship between the disk and the FAS group and the disks in the two groups.
  • the monitoring unit 22 is configured to monitor the update status of the disk, and when the disk is powered on or lose power, the power-on or power-off information is reported.
  • the new disk detecting unit 23 is configured to determine whether the powered disk is a new disk, and if so, trigger the new disk hot standby relationship determining unit 24, otherwise trigger the group determining unit 25.
  • the embodiment of the present invention focuses on the dynamic maintenance of the hot standby relationship between disks, in particular, the hot standby relationship maintenance when the disk is powered on. Therefore, the focus is on the new power-on of the disk.
  • the new disk hot standby relationship determining unit 24 is configured to determine a hot standby relationship of the new disk according to the group in which the new disk is located, and trigger a backup file loading unit 28, and the backup file loading unit 28 prepares the new disk for hot standby. The files in the object's disk are copied to the new disk.
  • the group judging unit 25 is configured to determine whether the current group of the power-on disk is consistent with the original group. If they are consistent, the hot standby relationship recovery and file content determining unit 26 is triggered. If not, the hot standby relationship updating unit 27 is triggered.
  • the hot standby relationship recovery and file content determining unit 26 is configured to restore the original hot standby relationship of the powered disk, and determine whether the file content of the powered disk is the same as the file content of the hot spare disk, and if not, the processing is not processed. If the process is different, the backup file loading unit 28 is triggered, and the backup file loading unit 28 updates the content of the file in the newly inserted disk and the hot spare disk to be consistent.
  • the hot standby relationship updating unit 27 is configured to re-establish a hot standby relationship for the powered disk according to the current group of the powered disk, and trigger the backup file loading unit 28, and the backup file loading unit 28 inserts the newly inserted disk.
  • the content of the file in the hot spare disk is updated to be consistent.
  • the update operation of the hot standby relationship is basically the same as the hot standby relationship update of the new disk insertion.
  • the backup file loading unit 28 is configured to load a backup file for the powered disk.
  • the apparatus for multi-disk group hot standby in the distributed file system of the embodiment of the present invention is designed to implement the method for multi-disk group hot standby in the foregoing distributed file system, and FIG. 2A and FIG. 2B
  • the implementation functions of the processing units in the display device can be understood by referring to the related description in the method shown in FIG. 1A.
  • the implementation functions of the processing units in the device shown in FIG. 2C can refer to the related description in the method shown in FIG. 1B.
  • the functions of the units in the multi-disk group hot standby device in the distributed file system of the embodiment of the present invention may be implemented by a program running on the processor, or may be implemented by a specific logic circuit.
  • the multi-disk group hot standby solution in the distributed file system provided by the embodiment of the present invention can make the hot standby relationship between the disks stronger, so that the service provided by the DFS to the user is more reliable, and will not be Easy to interrupt.
  • the implementation of the present invention does not modify the system architecture and the current processing flow, is easy to implement, facilitates promotion in the technical field, and has strong industrial applicability. The above is only the preferred embodiment of the present invention and is not intended to limit the scope of the present invention.

Description

分布式文件系统中多磁盘分组热备的方法与装置 技术领域 本发明涉及一种分布式文件系统( Distributed File System, 筒称为 DFS ) 中多磁盘分组热备的方法与装置。 背景技术
DFS中一^:包括多个文件访问月 务器( File Access Sever, 筒称为 FAS ), 每个 FAS对外提供文件系统的文件读写功能, 同时每个 FAS管理多个磁盘, 磁盘的犬态由 FAS定时向文件定位寄存器 (File Locating Register, 筒称为 FLR ) 上 4艮, 磁盘发现功能也由 FAS来实现。 现有技术中, 一般采用磁盘备 份的方法来提高系统的可靠性, 降低 FAS故障对整个系统产生的影响, 即, 在单个磁盘故障时, 使用一个备份的磁盘继续提供服务。 对于分布式文件系 统而言, 由于每个 FAS管理多个磁盘, 而每个 FAS管理的磁盘之间具有一 定的耦合性, 如果仅采用磁盘备份的机制, 当 FAS故障时, 该 FAS上的主 备盘将会同时故障, 从而导致该 FAS上的文件不能提供给访问用户, 造成系 统文件^^务的质量降氐。 发明内容 有鉴于此,本发明的主要目的在于提供一种分布式文件系统中多磁盘分 组热备的方法与装置, 能有效提高 DFS的可靠性。 为达到上述目的, 根据本发明的一个方面, 提供了一种分布式文件系统 中多磁盘分组热备的方法, 才艮据本发明的分布式文件系统中多磁盘分组热备 的方法包括: 以文件访问月 务器为对象对磁盘进行分组 ,以组为对象建立磁盘之间的 热备关系; 检测到磁盘上电时, 判断上电磁盘是否为新磁盘, 若是则根据所述新磁 盘所在的组确定所述新磁盘的热备关系, 为所述新磁盘加载备份文件; 若不是新磁盘则判断所述上电磁盘当前所在的组与原组是否一致 ,若一 致则恢复所述上电磁盘原有的热备关系, 并判断所述上电磁盘的文件内容与 其热备磁盘中文件内容是否相同, 若相同则不作处理, 若不同重新为所述上 电磁盘加载备份文件; 以及 若不一致则才艮据所述上电磁盘当前所在组重新为所述上电磁盘建立热 备关系, 并重新为所述上电磁盘加载备份文件。 优选地, 所述以文件访问月 务器为对象对磁盘进行分组, 具体为: 为所述文件访问^^务器配置位置的物理信息、逻辑单元信息以及分组号 信息, 建立磁盘参数与所述文件访问 ^^务器位置的物理信息、 逻辑单元信息 以及分组号信息之间的对应关系。 优选地, 所述位置的物理信息包括所述文件访问 ^^务器所在的架、 才匡、 槽板位的信息; 所述逻辑单元信息包括所述文件访问服务器的 CPU、 内口单 元、 夕卜口单元、 MAC地址、 IP地址的信息; 所述磁盘参数包括磁盘数据区 使用标识、磁盘在所述文件访问服务器的逻辑地址索引号、磁盘物理序列号、 磁盘逻辑标识、 磁盘块校验查询序列号、 磁盘心跳检测、 磁盘状态、 磁盘总 容量、 磁盘空闲容量、 磁盘某段时间读数据量、 磁盘某段时间写数据量、 上 •t艮开始时间、 上 4艮结束时间。 优选地, 所述以组为对象建立磁盘之间的热备关系, 具体为: 为所述组中的磁盘设置热备号,按所述热备号建立组之间磁盘的动态对 应关系。 根据本发明的另一个方面,提供了一种分布式文件系统中多磁盘分组热 备的装置。 根据本发明的分布式文件系统中多磁盘分组热备的装置包括: 磁盘分组单元, 用于以文件访问月 务器为对象对磁盘进行分组; 热备关系静态配置单元 , 用于以组为对象建立磁盘之间的热备关系; 监测单元, 用于监测磁盘的更新犬况; 新磁盘检测单元, 用于判断上电的磁盘是否为新磁盘, 若是则触发新磁 盘热备关系确定单元, 若否则触发组判断单元; 新磁盘热备关系确定单元,用于才艮据所述新磁盘所在的组确定所述新磁 盘的热备关系, 并触发备份文件加载单元; 组判断单元, 用于判断所述上电磁盘当前所在的组与原组是否一致, 若 一致则触发热备关系恢复及文件内容判断单元, 若不一致则触发热备关系更 新单元 热备关系恢复及文件内容判断单元,用于恢复所述上电磁盘原有的热备 关系, 并判断所述上电磁盘的文件内容与其热备磁盘中文件内容是否相同, 若相同则不作处理, 若不同则触发备份文件加载单元; 热备关系更新单元,用于才艮据所述上电磁盘当前所在组重新为所述上电 磁盘建立热备关系, 并触发备份文件加载单元; 以及 备份文件加载单元, 用于为所述上电磁盘加载备份文件。 优选地 , 所述磁盘分组单元以文件访问服务器为对象对磁盘进行分组 , 具体为: 为所述文件访问^^务器配置位置的物理信息、逻辑单元信息以及分组号 信息, 建立磁盘参数与所述文件访问 ^^务器位置的物理信息、 逻辑单元信息 以及分组号信息之间的对应关系。 优选地, 所述位置的物理信息包括所述文件访问 ^^务器所在的架、 才匡、 槽板位的信息; 所述逻辑单元信息包括所述文件访问服务器的 CPU、 内口单 元、 夕卜口单元、 MAC地址、 IP地址的信息; 所述磁盘参数包括磁盘数据区 使用标识、磁盘在所述文件访问服务器的逻辑地址索引号、磁盘物理序列号、 磁盘逻辑标识、 磁盘块校验查询序列号、 磁盘心跳检测、 磁盘状态、 磁盘总 容量、 磁盘空闲容量、 磁盘某段时间读数据量、 磁盘某段时间写数据量、 上 •t艮开始时间、 上 4艮结束时间。 优选地 ,所述热备关系静态配置单元以组为对象建立磁盘之间的热备关 系, 具体为: 为所述组中的磁盘设置热备号,按所述热备号建立组之间磁盘的动态对 应关系。 根据本发明的又一个方面 ,提供了一种分布式文件系统中多磁盘分组热 备的方法。 根据本发明的分布式文件系统中多磁盘分组热备的方法包括: 以 文件访问服务器为对象对磁盘分组 , 以组为对象建立磁盘之间的热备关系; 判断上电磁盘是否为新磁盘, 若是, 则才艮据所述新磁盘所在的组确定所述新 磁盘的热备关系, 为所述新磁盘加载备份文件, 否则, 才艮据所述上电磁盘当 前所在的组与原组是否一致配置所述上电磁盘的热备关系。 优选地, 在所述上电磁盘不是新磁盘的情况下, 根据所述上电磁盘当前 所在的组与原组是否一致配置所述上电磁盘的热备关系包括: 在上电磁盘当 前所在的组与原组一致的情况下, 恢复所述上电磁盘原有的热备关系, 并判 断所述上电磁盘的文件内容与其热备磁盘中文件内容是否相同 , 若相同则不 作处理, 若不同重新为所述上电磁盘加载备份文件; 在上电磁盘当前所在的 组与原组不一致的情况下 , 才艮据所述上电磁盘当前所在组重新为所述上电磁 盘建立热备关系, 并重新为所述上电磁盘加载备份文件。 根据本发明的再一个方面 ,提供了一种分布式文件系统中多磁盘分组热 备的装置。 根据本发明的分布式文件系统中多磁盘分组热备的装置包括: 预 配置单元, 用于以文件访问 ^^务器为对象对磁盘进行分组, 以组为对象建立 磁盘之间的热备关系; 新磁盘检测单元 , 用于判断上电磁盘是否为新磁盘 , 若是则触发新磁盘热备关系确定单元, 否则,触发旧磁盘热备关系配置单元; 所述新磁盘热备关系确定单元, 用于才艮据所述新磁盘所在的组确定所述新磁 盘的热备关系, 为所述新磁盘加载备份文件; 所述旧磁盘热备关系配置单元, 用于才艮据所述上电磁盘当前所在的组与原组是否一致配置所述上电磁盘的热 备关系。 优选地, 所述旧磁盘热备关系配置单元包括: 组判断子单元, 用于判断 所述上电磁盘当前所在的组与原组是否一致, 若一致则触发热备关系恢复子 单元, 若不一致则触发热备关系更新子单元; 所述热备关系恢复子单元, 用 于恢复所述上电磁盘原有的热备关系, 并判断所述上电磁盘的文件内容与其 热备磁盘中文件内容是否相同, 若相同则不作处理, 若不同则重新为所述上 电磁盘加载备份文件; 所述热备关系更新子单元, 用于才艮据所述上电磁盘当 前所在组重新为所述上电磁盘建立热备关系 , 并重新为所述上电磁盘加载备 份文件。 本发明以 FAS为磁盘分组的基础 ,一个 FAS中的所有磁盘划为一个组, 以组为单位进行热备 , 这样 , 无论是单个磁盘出现问题还是单个 FAS出现问 题, 都不会对整个 DFS造成影响。 另外, 在以组为单位进行热备时, 组内磁 盘之间的热备关系并不是固定的 , 而是根据磁盘的更新状况而动态变化的 , 当 FAS中有磁盘上电时,本发明能根据原先组之间的热备关系而自动生成新 上电磁盘的热备关系 , 从而使本发明磁盘间的热备关系更牢固。 本发明大大 提高了 DFS的数据的可靠性。 附图说明 此处所说明的附图用来提供对本发明的进一步理解 ,构成本申请的一部 分, 本发明的示意性实施例及其说明用于解释本发明, 并不构成对本发明的 不当限定。 在附图中: 图 1A为本发明实施例的分布式文件系统中多磁盘分组热备的方法的流 程示意图; 图 1B为本发明实施例的分布式文件系统中多磁盘分组热备的方法的详 细¾¾程示意图; 图 2A为本发明实施例的分布式文件系统中多磁盘分组设备的装置的结 构框图; 图 2B为本发明实施例的分布式文件系统中多磁盘分组设备的装置中旧 磁盘热备关系配置单元的详细结构才匡图; 图 2C为本发明实施例的分布式文件系统中多磁盘分组热备的装置的详 细结构框图。 具体实施方式 功能相无述 本发明的基本思想是: 以 FAS为磁盘分组的基础, 一个 FAS中的所有 磁盘划为一个组, 以组为单位进行热备, 这样, 无论是单个磁盘出现问题还 是单个 FAS 出现问题, 都不会对整个 DFS造成影响。 另外, 在以组为单位 进行热备时, 组内磁盘之间的热备关系并不是固定的, 而是根据磁盘的更新 状况而动态变化的, 当 FAS中有磁盘上电时, 能够才艮据原先组之间的热备关 系而自动生成新上电磁盘的热备关系 , 从而使磁盘间的热备关系更牢固 , 使 DFS向用户提供数据文件的服务更可靠, 不会轻易中断。 为使本发明的目的、技术方案和优点更加清楚明白, 以下举实施例并参 照附图 , 对本发明进一步详细说明。 方法实施例 图 1A为本发明实施例的分布式文件系统中多磁盘分组热备的方法的流 程示意图, 如图 1A所示, 该方法包括: 步骤 101a, 以文件访问服务器为对象对磁盘分组, 以组为对象建立磁 盘之间的热备关系; 步骤 102a, 判断上电磁盘是否为新磁盘, 若是, 则才艮据新磁盘所在的 组确定新磁盘的热备关系, 为新磁盘加载备份文件, 否则, 才艮据上电磁盘当 前所在的组与原组是否一致配置上电磁盘的热备关系。 在步骤 102a中, 在上电磁盘不是新磁盘的情况下, 根据上电磁盘当前 所在的组与原组是否一致配置上电磁盘的热备关系包括: 在上电磁盘当前所 在的组与原组一致的情况下, 恢复上电磁盘原有的热备关系, 并判断上电磁 盘的文件内容与其热备磁盘中文件内容是否相同, 若相同则不作处理, 若不 同重新为上电磁盘加载备份文件; 在上电磁盘当前所在的组与原组不一致的 情况下, 才艮据上电磁盘当前所在组重新为上电磁盘建立热备关系, 并重新为 上电磁盘加载备份文件。 图 1B为本发明实施例的分布式文件系统中多磁盘分组热备的方法的详 细流程示意图, 如图 1B所示, 该方法包括: 步骤 101b: 以文件访问月 务器为对象对磁盘进行分组, 以组为对象建 立磁盘之间的热备关系。 在 FLR上留出存放磁盘信息的数据区 , 这里的磁盘信息可以包括磁盘 的基本信息 , 磁盘所属 FAS的信息 , 组中磁盘之间的热备关系等。 FLR的数 据区为 FAS配置位置的物理信息、 逻辑单元信息以及分组号信息, 建立磁盘 参数与 FAS 位置的物理信息、 逻辑单元信息以及分组号信息之间的对应关 系。 所述位置的物理信息可以包括所述文件访问服务器所在的架、 框、 槽板 位的信息等;所述逻辑单元信息可以包括 FAS的 CPU、 内口单元、外口单元、 MAC地址、 IP地址的信息等; 所述磁盘参数可以包括磁盘数据区使用标识、 磁盘在所述文件访问服务器的逻辑地址索引号、 磁盘物理序列号、 磁盘逻辑 标识、 磁盘块校验查询序列号、 磁盘心跳检测、 磁盘状态、 磁盘总容量、 磁 盘空闲容量、 磁盘某段时间读数据量、 磁盘某段时间写数据量、 上 4艮开始时 间、 上 ·ί艮结束时间等。 本发明实施例以一个 FAS 中的所有磁盘为一个组, 在两个组之间建立 起热备关系, 为所述组中的磁盘设置热备号, 按所述热备号建立组之间磁盘 的动态对应关系。 虽然两个组之间是热备关系 , 但组内的磁盘之间的热备对 应关系却是动态的: 每个组内热备号码数按 FAS支持的最大磁盘数来设置, 以在 FAS中的磁盘上电时,仍能按本发明实施例提供的热备方法进行动态调 整。 本发明实施例中, 互为热备的两个组分配相同的组号, 以便于只别。 也 就是说, 本发明实施例将热备关系限定到组, 组内磁盘之间的热备关系可任 意设定。 步骤 102b: 检测到磁盘上电时, 判断上电磁盘是否为新磁盘, 若是则 执行步骤 103b, 否则执行步骤 104b。 本发明实施例中, FAS会周期性地检测该 FAS 中的磁盘更新状况, 当 有磁盘插入上电或磁盘被拔掉或坏掉时, 均会将检测结果上 4艮至 FLR, FLR 根据 FAS的上报结果对 FLR数据区进行修改, 增加或删除磁盘的相关信息 等操作。 由于本发明实施例针对的是磁盘的动态热备方法, 因此将重点描述 磁盘上电的情况, 对于磁盘脱离 FLR的情况, 磁盘的相关配置信息会保留一 段时间, 超出设定预设的时长后, 删除磁盘的相关配置信息。 当有磁盘插入 到 FAS中的磁盘插口时 , FAS动态检测磁盘的相关参数信息 , 包括磁盘物理 序列号、 磁盘总容量、 磁盘空闲容量等, 具体的参数信息参见步骤 101b 中 的相关描述, 这里不再赞述; FAS上报所检测到的磁盘的参数信息。 FLR根 据所接收到的磁盘参数信息检索数据区, 查看所述磁盘是否为新磁盘, 若是 则执行步骤 103b, 否则执行步骤 104b。 步骤 103b: 才艮据新磁盘所在的组确定新磁盘的热备关系, 并为新磁盘 加载备份文件。
FLR根据上报磁盘参数信息的 FAS确定该新磁盘所属的组, 根据所述 组信息为新磁盘分配一个热备号码, 并确定与该组互为热备的另一个组, 如 前所述, 可通过组号来确定互为热备的两个组, 确定所述另一个组中原来没 有热备的磁盘, 如果有两个, 则任选一个所述原没有热备的磁盘与新磁盘建 立互为热备关系, 如果有一个, 则将该原没有热备的磁盘与新磁盘建立互为 热备关系, 如果没有则不作处理; 其中, 建立热备关系即将所述原没有热备 的磁盘与新磁盘之间的热备号码关联起来, 并存储于 FLR的数据区。 建立起 热备关系后,将所述原没有热备的磁盘中的文件拷贝到新磁盘中, 结束流程。 步骤 104b: 判断所述上电磁盘当前所在的组与原组是否一致, 若一致 则执行步骤 105b, 否则执行步骤 106b。
FLR才艮据磁盘的相关参数信息确定所述上电磁盘原来所属的组,确定所 述上电磁盘是否更换了 FAS , 如果更换了 FAS, 则该上电磁盘是从其它 FAS 中更换来的, 执行步骤 106b, 否则该上电磁盘是 FAS 内磁盘重新上电使用 的, 执行步骤 105b。 步骤 105b: 恢复磁盘间原有的热备关系, 并判断所述上电磁盘的文件 内容与其热备磁盘中文件内容是否相同, 若相同则不作处理, 若不同则重新 为所述上电磁盘加载备份文件。
FAS 中插入原有的磁盘, 很有可能是 FAS 中磁盘因接口虚接、 短暂维 修等原因而暂时取下而又插入到 FAS中的, 此时, FLR查找该磁盘的热备关 系是否还存在, 存在时查看该磁盘中的文件内容是否与热备磁盘中的文件内 容一致, 一致时不作处理, 否则进行文件更新, 使所述新插入磁盘与热备磁 盘中的文件内容一致。 如果不存在所述磁盘的热备关系, 则恢复所述上电磁 盘原有的热备关系, 再查看该磁盘中的文件内容是否与热备磁盘中的文件内 容一致, 一致时不作处理, 否则^)夺所述新插入磁盘与热备磁盘中的文件内容 更新为一致。 结束流程。 步骤 106b: 才艮据所述上电磁盘当前所在组重新为所述上电磁盘建立热 备关系, 并重新为所述上电磁盘加载备份文件。 当确定新插入 FAS中的磁盘原来不属于该 FAS时 , 将该新插入的磁盘 当作新磁盘进行处理, 即, 按步骤 103b 的方式进行热备处理, 这里不再赘 述。 磁盘热备关系处理完毕后, 将所述新插入磁盘与热备磁盘中的文件内容 更新为一致, 删除所述上电磁盘的原来的热备配置信息。 结束流程。 才艮据本发明实施例, 还提供了一种计算机可读介质, 该计算机可读介质 上存储有计算机可执行的指令, 当该指令被计算机或处理器执行时, 使得计 算机或处理器执行如图 1A、 图 1B所示的各步骤的处理, 优选地, 可以执行 上述实施例中的一个或多个。 装置实施例 图 2A为本发明实施例的分布式文件系统中多磁盘分组设备的装置的结 构框图, 如图 2A所示, 本发明实施例的分布式文件系统中多磁盘分组热备 的装置包括:预配置单元 12,用于以文件访问 ^^务器为对象对磁盘进行分组, 以组为对象建立磁盘之间的热备关系; 新磁盘检测单元 14, 连接于预配置单 元 12 , 用于判断上电磁盘是否为新磁盘, 若是则触发新磁盘热备关系确定单 元 16, 否则, 触发旧磁盘热备关系配置单元 18; 新磁盘热备关系确定单元 16, 连接于新磁盘检测单元 14, 用于才艮据新磁盘所在的组确定新磁盘的热备 关系, 为新磁盘加载备份文件; 旧磁盘热备关系配置单元 18 , 连接于新磁盘 检测单元 14 , 用于才艮据上电磁盘当前所在的组与原组是否一致配置上电磁盘 的热备关系。 图 2B为本发明实施例的分布式文件系统中多磁盘分组设备的装置中旧 磁盘热备关系配置单元的详细结构框图, 如图 2B 所示, 旧磁盘热备关系配 置单元 18包括: 组判断子单元 182 , 用于判断上电磁盘当前所在的组与原组 是否一致, 若一致则触发热备关系恢复子单元, 若不一致则触发热备关系更 新子单元; 热备关系恢复子单元 184, 连接于组判断子单元 182 , 用于恢复 上电磁盘原有的热备关系, 并判断上电磁盘的文件内容与其热备磁盘中文件 内容是否相同, 若相同则不作处理, 若不同则重新为上电磁盘加载备份文件; 热备关系更新子单元 186, 连接于组判断子单元 182, 用于根据上电磁盘当 前所在组重新为上电磁盘建立热备关系, 并重新为上电磁盘加载备份文件。 图 2C为本发明实施例的分布式文件系统中多磁盘分组热备的装置的详 细结构框图, 如图 2C所示, 本发明实施例的分布式文件系统中多磁盘分组 热备的装置包括磁盘分组单元 20、 热备关系静态配置单元 21、 监测单元 22、 新磁盘检测单元 23、 新磁盘热备关系确定单元 24、 组判断单元 25、 热备关 系恢复及文件内容判断单元 26、 热备关系更新单元 27和备份文件加载单元 28, 其中, 磁盘分组单元 20用于以文件访问服务器为对象对磁盘进行分组, 即 , 一个 FAS中的所有磁盘归为一个组。 热备关系静态配置单元 21用于以 组为对象建立磁盘之间的热备关系, 即, 建立组之间的热备关系, 但并不限 定组中各磁盘之间的热备对应关系, 具有热备关系的组选用相同的组号。 关 于热备关系的建立,需要上报磁盘的基本参数信息以及 FAS相关信息,其中, 热备关系即磁盘与 FAS组、 两组中磁盘之间的关联关系。 监测单元 22用于 监测磁盘的更新状况, 当磁盘上电或失电时, 将上电或失电信息进行上报。 新磁盘检测单元 23 用于判断上电的磁盘是否为新磁盘, 若是则触发新磁盘 热备关系确定单元 24 , 否则触发组判断单元 25。 本发明实施例的重点是磁 盘间热备关系的动态维护, 特别是有磁盘上电时的热备关系维护, 因此重点 关注磁盘新上电的情形。 新磁盘热备关系确定单元 24 用于才艮据所述新磁盘 所在的组确定所述新磁盘的热备关系, 并触发备份文件加载单元 28, 备份文 件加载单元 28 将所述新磁盘热备对象的磁盘中的文件拷贝到新磁盘中。 组 判断单元 25 用于判断所述上电磁盘当前所在的组与原组是否一致, 若一致 则触发热备关系恢复及文件内容判断单元 26, 若不一致则触发热备关系更新 单元 27。 热备关系恢复及文件内容判断单元 26用于恢复所述上电磁盘原有 的热备关系, 并判断所述上电磁盘的文件内容与其热备磁盘中文件内容是否 相同, 若相同则不作处理, 结束流程; 若不同则触发备份文件加载单元 28, 备份文件加载单元 28 将所述新插入磁盘与热备磁盘中的文件内容更新为一 致。 热备关系更新单元 27 用于才艮据所述上电磁盘当前所在组重新为所述上 电磁盘建立热备关系, 并触发备份文件加载单元 28 , 备份文件加载单元 28 将所述新插入磁盘与热备磁盘中的文件内容更新为一致。 当原有磁盘是其他 FAS中的磁盘而新插入了另一个 FAS中时 ,热备关系的更新动作与新磁盘插 入的热备关系更新基本相同。 备份文件加载单元 28 用于为所述上电磁盘加 载备份文件。 本领域技术人员应当理解,本发明实施例的分布式文件系统中多磁盘分 组热备的装置是为实现前述分布式文件系统中多磁盘分组热备的方法而设计 的, 图 2A、 图 2B所示装置中的各处理单元的实现功能可参照图 1A所示的 方法中的相关描述而理解, 图 2C所示装置中的各处理单元的实现功能可参 照图 1B所示的方法中的相关描述而理解。 本发明实施例的分布式文件系统 中多磁盘分组热备的装置中各单元的功能可通过运行于处理器上的程序而实 现, 也可通过具体的逻辑电路而实现。 综上所述,通过本发明实施例提供的分布式文件系统中多磁盘分组热备 的方案, 能够使磁盘间的热备关系更牢固 , 使 DFS向用户提供数据文件的服 务更可靠, 不会轻易中断。 另外 ,本发明的实现没有对系统架构和目前的处理流程修改,易于实现, 便于在技术领域中进行推广, 具有较强的工业适用性。 以上所述, 仅为本发明的较佳实施例而已, 并非用于限定本发明的保护 范围。

Claims

权 利 要 求 书
1. 一种分布式文件系统中多磁盘分组热备的方法, 其特征在于, 包括: 以文件访问月 务器为对象对磁盘进行分组 , 以组为对象建立磁盘 之间的热备关系;
检测到磁盘上电时, 判断上电磁盘是否为新磁盘, 若是则才艮据所 述新磁盘所在的组确定所述新磁盘的热备关系 , 为所述新磁盘加载备 份文件;
若不是新磁盘则判断所述上电磁盘当前所在的组与原组是否一 致, 若一致则恢复所述上电磁盘原有的热备关系, 并判断所述上电磁 盘的文件内容与其热备磁盘中文件内容是否相同 ,若相同则不作处理 , 若不同重新为所述上电磁盘加载备份文件; 以及
若不一致则才艮据所述上电磁盘当前所在组重新为所述上电磁盘 建立热备关系, 并重新为所述上电磁盘加载备份文件。
2. 根据权利要求 1所述的方法, 其特征在于, 所述以文件访问服务器为 对象对磁盘进行分组, 具体为:
为所述文件访问服务器配置位置的物理信息、 逻辑单元信息以及 分组号信息, 建立磁盘参数与所述文件访问服务器位置的物理信息、 逻辑单元信息以及分组号信息之间的对应关系。
3. 根据权利要求 2所述的方法, 其特征在于, 所述位置的物理信息包括 所述文件访问服务器所在的架、 框、 槽板位的信息; 所述逻辑单元信 息包括所述文件访问月 务器的 CPU、 内口单元、 夕卜口单元、 MAC 地 址、 IP地址的信息; 所述磁盘参数包括磁盘数据区使用标识、 磁盘在 所述文件访问服务器的逻辑地址索引号、 磁盘物理序列号、 磁盘逻辑 标识、 磁盘块校验查询序列号、 磁盘心兆检测、 磁盘状态、 磁盘总容 量、 磁盘空闲容量、 磁盘某段时间读数据量、 磁盘某段时间写数据量、 上 4艮开始时间、 上 4艮结束时间。 根据权利要求 1所述的方法, 其特征在于, 所述以组为对象建立磁盘 之间的热备关系, 具体为:
为所述组中的磁盘设置热备号 , 按所述热备号建立组之间磁盘的 动态对应关系。 一种分布式文件系统中多磁盘分组热备的装置, 其特征在于, 包括: 磁盘分组单元 , 用于以文件访问 务器为对象对磁盘进行分组; 热备关系静态配置单元, 用于以组为对象建立磁盘之间的热备关 系;
监测单元 , 用于监测磁盘的更新状况;
新磁盘检测单元, 用于判断上电的磁盘是否为新磁盘 , 若是则触 发新磁盘热备关系确定单元, 若否则触发组判断单元;
新磁盘热备关系确定单元 , 用于才艮据所述新磁盘所在的组确定所 述新磁盘的热备关系, 并触发备份文件加载单元;
组判断单元, 用于判断所述上电磁盘当前所在的组与原组是否一 致, 若一致则触发热备关系恢复及文件内容判断单元, 若不一致则触 发热备关系更新单元;
热备关系恢复及文件内容判断单元, 用于恢复所述上电磁盘原有 的热备关系, 并判断所述上电磁盘的文件内容与其热备磁盘中文件内 容是否相同 , 若相同则不作处理 , 若不同则触发备份文件加载单元; 热备关系更新单元, 用于才艮据所述上电磁盘当前所在组重新为所 述上电磁盘建立热备关系, 并触发备份文件加载单元; 以及
备份文件加载单元, 用于为所述上电磁盘加载备份文件。 根据权利要求 5 中所述的装置, 其特征在于, 所述磁盘分组单元以文 件访问月 务器为对象对磁盘进行分组 , 具体为:
为所述文件访问服务器配置位置的物理信息、 逻辑单元信息以及 分组号信息, 建立磁盘参数与所述文件访问服务器位置的物理信息、 逻辑单元信息以及分组号信息之间的对应关系。
7. 根据权利要求 6所述的装置, 其特征在于, 所述位置的物理信息包括 所述文件访问服务器所在的架、 框、 槽板位的信息; 所述逻辑单元信 息包括所述文件访问月 务器的 CPU、 内口单元、 夕卜口单元、 MAC 地 址、 IP地址的信息; 所述磁盘参数包括磁盘数据区使用标识、 磁盘在 所述文件访问服务器的逻辑地址索引号、 磁盘物理序列号、 磁盘逻辑 标识、 磁盘块校验查询序列号、 磁盘心兆检测、 磁盘状态、 磁盘总容 量、 磁盘空闲容量、 磁盘某段时间读数据量、 磁盘某段时间写数据量、 上 4艮开始时间、 上 4艮结束时间。
8. 根据权利要求 5所述的装置, 其特征在于, 所述热备关系静态配置单 元以组为对象建立磁盘之间的热备关系 , 具体为:
为所述组中的磁盘设置热备号 , 按所述热备号建立组之间磁盘的 动态对应关系。
9. 一种分布式文件系统中多磁盘分组热备的方法, 其特征在于, 包括: 以文件访问月 务器为对象对磁盘分组 , 以组为对象建立磁盘之间 的热备关系;
判断上电磁盘是否为新磁盘, 若是, 则根据所述新磁盘所在的组 确定所述新磁盘的热备关系, 为所述新磁盘加载备份文件, 否则, 根 据所述上电磁盘当前所在的组与原组是否一致配置所述上电磁盘的热 备关系。
10. 根据权利要求 9所述的方法, 其特征在于, 在所述上电磁盘不是新磁 盘的情况下 , 才艮据所述上电磁盘当前所在的组与原组是否一致配置所 述上电磁盘的热备关系包括:
在上电磁盘当前所在的组与原组一致的情况下 , 恢复所述上电磁 盘原有的热备关系, 并判断所述上电磁盘的文件内容与其热备磁盘中 文件内容是否相同, 若相同则不作处理, 若不同重新为所述上电磁盘 加载备份文件;
在上电磁盘当前所在的组与原组不一致的情况下 , 才艮据所述上电 磁盘当前所在组重新为所述上电磁盘建立热备关系, 并重新为所述上 电磁盘加载备份文件。
11. 一种分布式文件系统中多磁盘分组热备的装置, 其特征在于, 包括: 预配置单元, 用于以文件访问 务器为对象对磁盘进行分组 , 以 组为对象建立磁盘之间的热备关系;
新磁盘检测单元, 用于判断上电磁盘是否为新磁盘, 若是则触发 新磁盘热备关系确定单元, 否则, 触发旧磁盘热备关系配置单元; 所述新磁盘热备关系确定单元 , 用于才艮据所述新磁盘所在的组确 定所述新磁盘的热备关系 , 为所述新磁盘加载备份文件;
所述旧磁盘热备关系配置单元 , 用于才艮据所述上电磁盘当前所在 的组与原组是否一致配置所述上电磁盘的热备关系。 根据权利要求 11所述的装置, 其特征在于, 所述旧磁盘热备关系配置 单元包括:
组判断子单元, 用于判断所述上电磁盘当前所在的组与原组是否 一致, 若一致则触发热备关系恢复子单元, 若不一致则触发热备关系 更新子单元; 所述热备关系恢复子单元, 用于恢复所述上电磁盘原有的热备关 系, 并判断所述上电磁盘的文件内容与其热备磁盘中文件内容是否相 同 , 若相同则不作处理 , 若不同则重新为所述上电磁盘加载备份文件; 所述热备关系更新子单元, 用于才艮据所述上电磁盘当前所在组重 新为所述上电磁盘建立热备关系, 并重新为所述上电磁盘加载备份文 件。
PCT/CN2009/073866 2008-12-31 2009-09-10 分布式文件系统中多磁盘分组热备的方法与装置 WO2010075703A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
EP09835990.4A EP2378434B1 (en) 2008-12-31 2009-09-10 Mehtod and device of multiple disks grouping hot standby in distributed file system

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN2008102468078A CN101441668B (zh) 2008-12-31 2008-12-31 分布式文件系统中多磁盘分组热备的方法与装置
CN200810246807.8 2008-12-31

Publications (1)

Publication Number Publication Date
WO2010075703A1 true WO2010075703A1 (zh) 2010-07-08

Family

ID=40726102

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2009/073866 WO2010075703A1 (zh) 2008-12-31 2009-09-10 分布式文件系统中多磁盘分组热备的方法与装置

Country Status (3)

Country Link
EP (1) EP2378434B1 (zh)
CN (1) CN101441668B (zh)
WO (1) WO2010075703A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2016070537A1 (zh) * 2014-11-04 2016-05-12 中兴通讯股份有限公司 一种磁盘有效性的检测方法、认证方法及相应装置

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101441668B (zh) * 2008-12-31 2011-09-21 中兴通讯股份有限公司 分布式文件系统中多磁盘分组热备的方法与装置
CN102053801A (zh) * 2010-12-29 2011-05-11 成都市华为赛门铁克科技有限公司 一种磁盘热备方法及装置、存储系统
CN103064757A (zh) * 2012-12-12 2013-04-24 鸿富锦精密工业(深圳)有限公司 数据备份方法及系统
CN105159847A (zh) * 2015-08-12 2015-12-16 北京因特信安软件科技有限公司 基于可信芯片的磁盘变更记录方法
CN109739340B (zh) * 2017-08-15 2021-04-13 金钱猫科技股份有限公司 一种节能省电的数据处理方法及终端

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN100341255C (zh) * 2002-12-26 2007-10-03 华为技术有限公司 一种交叉主备保护的热备份方法
CN100363900C (zh) * 2003-05-19 2008-01-23 日立环球储存科技荷兰有限公司 Raid-1系统中的备用系统和方法
WO2008036319A2 (en) * 2006-09-18 2008-03-27 Lsi Logic Optimized reconstruction and copyback methodology for a disconnected drive in the presence of a global hot spare disk
CN101441668A (zh) * 2008-12-31 2009-05-27 中兴通讯股份有限公司 分布式文件系统中多磁盘分组热备的方法与装置

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6405284B1 (en) * 1998-10-23 2002-06-11 Oracle Corporation Distributing data across multiple data storage devices in a data storage system
US6862692B2 (en) * 2001-01-29 2005-03-01 Adaptec, Inc. Dynamic redistribution of parity groups

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN100341255C (zh) * 2002-12-26 2007-10-03 华为技术有限公司 一种交叉主备保护的热备份方法
CN100363900C (zh) * 2003-05-19 2008-01-23 日立环球储存科技荷兰有限公司 Raid-1系统中的备用系统和方法
WO2008036319A2 (en) * 2006-09-18 2008-03-27 Lsi Logic Optimized reconstruction and copyback methodology for a disconnected drive in the presence of a global hot spare disk
CN101441668A (zh) * 2008-12-31 2009-05-27 中兴通讯股份有限公司 分布式文件系统中多磁盘分组热备的方法与装置

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2016070537A1 (zh) * 2014-11-04 2016-05-12 中兴通讯股份有限公司 一种磁盘有效性的检测方法、认证方法及相应装置

Also Published As

Publication number Publication date
EP2378434B1 (en) 2014-09-03
EP2378434A4 (en) 2013-03-13
EP2378434A1 (en) 2011-10-19
CN101441668A (zh) 2009-05-27
CN101441668B (zh) 2011-09-21

Similar Documents

Publication Publication Date Title
US11487632B2 (en) Techniques for LIF placement in SAN storage cluster synchronous disaster recovery
CN110071821B (zh) 确定事务日志的状态的方法,节点和存储介质
CN104715001B (zh) 用于对数据处理系统的集群中的共享资源执行写入操作的方法和系统
EP2273371B1 (en) Failover procedure for server system
US8185961B2 (en) Network system, method for controlling access to storage device, management server, storage device, log-in control method, network boot system, and method of accessing individual storage unit
JP4349871B2 (ja) ファイル共有装置及びファイル共有装置間のデータ移行方法
US10693955B2 (en) Techniques for SAN storage cluster synchronous disaster recovery
US20190108231A1 (en) Application Aware Snapshots
JP4572250B2 (ja) 計算機切り替え方法、計算機切り替えプログラム及び計算機システム
JP4939102B2 (ja) ネットワークブート計算機システムの高信頼化方法
US7447933B2 (en) Fail-over storage system
JP4718288B2 (ja) ディスクレス計算機の運用管理システム
US8046632B2 (en) Backup management method based on mode of failure
WO2013037235A1 (zh) 实现用户系统还原的方法、系统和装置及存储介质
WO2010075703A1 (zh) 分布式文件系统中多磁盘分组热备的方法与装置
US20100169470A1 (en) System and method for operational management of computer system
US8380951B1 (en) Dynamically updating backup configuration information for a storage cluster
JP2005267327A (ja) ストレージシステム
JP2010108063A (ja) 監視対象装置管理システム、管理サーバおよび監視対象装置管理方法
US20070192553A1 (en) Backup apparatus and backup method
US11275599B1 (en) Hypervisor remote offload for VM data migration between storage devices
JP2008065392A (ja) ストレージ装置及びこれを用いたデータのバックアップ方法
WO2012000272A1 (zh) 一种文件系统被破坏后的自救方法和设备
WO2018196154A1 (zh) 任务处理方法、系统、终端及计算机可读存储介质
JP5484434B2 (ja) ネットワークブート計算機システム、管理計算機、及び計算機システムの制御方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 09835990

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

WWE Wipo information: entry into national phase

Ref document number: 2009835990

Country of ref document: EP

WWE Wipo information: entry into national phase

Ref document number: 4932/CHENP/2011

Country of ref document: IN