WO2008148319A1 - Procédé de récupération d'image, équipement de stockage et système réseau - Google Patents

Procédé de récupération d'image, équipement de stockage et système réseau Download PDF

Info

Publication number
WO2008148319A1
WO2008148319A1 PCT/CN2008/070879 CN2008070879W WO2008148319A1 WO 2008148319 A1 WO2008148319 A1 WO 2008148319A1 CN 2008070879 W CN2008070879 W CN 2008070879W WO 2008148319 A1 WO2008148319 A1 WO 2008148319A1
Authority
WO
WIPO (PCT)
Prior art keywords
storage device
location
data
location record
record
Prior art date
Application number
PCT/CN2008/070879
Other languages
English (en)
French (fr)
Inventor
Peng Zhang
Guobin Zhang
Ji Ouyang
Sheng Zhou
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.
Publication of WO2008148319A1 publication Critical patent/WO2008148319A1/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
    • G06F11/2056Error 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 by mirroring
    • G06F11/2082Data synchronisation

Definitions

  • the present invention relates to the field of computer network technologies, and in particular, to a image restoration method, a storage device, and a network system.
  • Remote mirroring is one of the technologies for building a disaster recovery backup system.
  • Mirroring is the process of storing information that produces a mirrored view of the same data on two or more storage devices.
  • Remote mirroring can be divided into synchronous remote mirroring and asynchronous remote mirroring.
  • Synchronous remote mirroring (synchronous replication technology) means that each time the application host of the primary station issues a write request, it must wait until the data is completely written to the primary volume of the primary storage device and the secondary volume of the secondary storage device before continuing. Synchronous remote mirroring is generally suitable for short communication links (usually within a few tens of kilometers).
  • Asynchronous remote mirroring (asynchronous replication technology), when the application host of the primary station continuously sends multiple write requests, it only needs to wait for the data to be written to the primary volume of the primary storage device to continue execution, and then the primary volume of the primary storage device. Multiple data synchronizations are continued with the slave volume from the storage device, ie multiple operations are allowed to lag between the two storage devices.
  • Asynchronous remote mirroring is useful when the communication link is long (usually hundreds or even thousands of miles).
  • the asynchronous remote mirroring process involves mirror recovery issues.
  • the update data on the primary volume is periodically sent to the slave for updates.
  • the primary volume temporarily expires, and the secondary volume becomes the new primary volume.
  • the data of the new primary volume is completely copied to the original primary volume, and the mirror recovery is completed.
  • FIG. 1 is a flowchart of the prior art asynchronous remote mirror recovery.
  • the recovery technology of the existing asynchronous remote mirror is that after the primary volume fails, the service from the volume of the primary volume becomes the new primary volume. After a period of time, the primary volume is restored and the business needs to switch back to the original primary volume from the new primary volume. At this point, the data for the new primary volume is completely copied to the original primary volume, and then the business is switched back to the original primary volume.
  • Step A1 the primary volume sends update data to the slave volume for update;
  • the application host sends a write request to the primary volume of the primary storage device.
  • the asynchronous remote mirror program of the primary storage device writes the update data of the write request to the primary volume of the primary storage device according to the location requested in the write request.
  • the identifier of the corresponding position in the bitmap is set to "1", indicating that the data of the location has been updated. If the identifier of a certain location in the bitmap is "0", it indicates that the data of the location is not updated.
  • a bitmap is also a sequence of binary sequences. Each bit in the sequence is "0" or "1", indicating two different states.
  • the update data of the corresponding location is sent to the slave volume according to the information recorded in the bitmap, and is updated after being received from the volume.
  • Step A2 The primary volume is temporarily invalidated, and the new volume is changed from the volume takeover service
  • the clustered software of the application host cooperates with the asynchronous remote mirroring program of the master/slave storage device, and the service of the primary volume is switched to take over from the volume, and the secondary volume becomes the new primary volume.
  • bitmap is also used to record where the data update occurred.
  • Step A3 The original primary volume is restored, and all the data of the new primary volume is copied to the original primary volume.
  • the storage device periodically detects whether the original primary volume of the primary storage device is normal. If the original primary volume is detected to be normal, the original primary volume is considered to have been restored. After the original primary volume is restored, all the data of the new primary volume is copied to the original primary volume, that is, the new primary volume sends all the data to the original primary volume, and the original primary volume is updated after receiving, thereby completing the image recovery. Then, after the cluster software of the application host cooperates with the asynchronous remote mirror program of the master/slave storage device, the service is switched from the new primary volume to the original primary volume.
  • the inventor found that the image recovery speed is relatively slow and takes a long time to meet the requirements of the disaster tolerance level.
  • the technical problem to be solved by the embodiments of the present invention is the problem that the image recovery speed in the existing image recovery method, the storage device, and the network system is relatively slow.
  • the embodiment of the present invention provides a mirror image recovery method, including: acquiring a location record of a data update by a first storage device, where the location record records each location The update status of the data; the second storage device compares the location record of the local end with the acquired location record of the first storage device, obtains a location that contains different data from the first storage device; and copies the data corresponding to the location to First storage Ready.
  • An embodiment of the present invention provides a storage device, including: an acquiring unit, configured to acquire a location record of data update by another storage device, where the location record records an update status of data at each location; and a recording unit, configured to record the a location record of the local end of the storage device; a comparing unit, configured to compare a location record of the local end recorded by the recording unit with a location record of the another storage device acquired by the acquiring unit, to obtain the other storage device a location containing different data; a data sending unit, configured to send data corresponding to the location obtained by the comparing unit to the another storage device.
  • An embodiment of the present invention provides a network system, including: a first storage device, configured to send a location record for performing data update, receive data and update, and the location record records an update status of data at each location; the second storage device And acquiring a location record sent by the first storage device, comparing the location record of the local end with the acquired location record of the first storage device, and obtaining a location that contains different data from the first storage device, where The data corresponding to the location is sent to the first storage device.
  • the technical solution provided by the embodiment of the present invention is: obtaining a location record of data update by the first storage device, where the location record records an update status of data at each location; and the second storage device locates the local end Recording is compared with the acquired location record of the first storage device to obtain a location containing different data from the first storage device; and copying the data corresponding to the location to the first storage device.
  • the second storage device and the first storage device are obtained by adding a process of comparing the location record of the data update of the second storage device with the location record of the data update performed by the first storage device.
  • the data corresponding to these locations is the data that needs to be copied to the first storage device to complete the image recovery, so that it is not necessary to copy all the data as in the prior art, thereby improving the image recovery speed.
  • the asynchronous remote mirroring can be applied to the environment with high disaster tolerance level, and the data to be copied is reduced, the burden of the link is reduced, and the fault tolerance of the system is improved.
  • FIG. 2 is a schematic diagram of an application environment of a mirror recovery method according to an embodiment of the present invention
  • FIG. 3 is a flowchart of a mirror recovery method according to an embodiment of the present invention.
  • FIG. 4 is a flowchart of a method for image restoration in the second embodiment of the present invention
  • FIG. 5 is a schematic structural diagram of a storage device according to an embodiment of the present invention
  • FIG. 6 is a schematic structural diagram of a network system according to an embodiment of the present invention.
  • the image recovery speed of the existing image recovery method is relatively slow.
  • the embodiment of the present invention provides a mirror recovery method, which can improve the image recovery speed.
  • FIG. 2 is a schematic diagram of an application environment of a mirror recovery method according to an embodiment of the present invention.
  • Figure 2 shows one of the networking scenarios of asynchronous remote mirroring, including two application hosts, which are interconnected by a wide area network (WAN) or a local area network (LAN) to install clusters on two application hosts.
  • the software forms an environment of highly available test HA (High Availability).
  • the services running on the application host can be enterprise applications such as database services, mail services, and file services.
  • the networking also includes two storage devices, the primary storage device and the secondary storage device.
  • the service data of the application host is saved in the primary volume of the primary storage device and the secondary volume of the secondary storage device.
  • the application host is interconnected with the primary and secondary storage devices through a storage area network (SAN).
  • SAN storage area network
  • FC protocol Fibre Channel, Fibre Channel
  • iSCSI Internet SCSI, Internet SCSI
  • SAS protocol Attached SCSI, Serial SCSI technology, etc.
  • SCSI is the English abbreviation of Small Computer System Interface, which is generally called the small computer system interface.
  • the service data of the application host is first written to the primary volume of the primary storage device, and then the primary volume then mirrors the data to the secondary volume of the secondary storage device.
  • the data link between the primary storage device and the secondary storage device may be FC (Fibre Channel, Fibre Channel), DWDM (Dense Wave Length Division Multiplexing), ATM (Asynchronous Transfer Mode), or IP ( Transmission lines such as Internet Protocol, Internet Protocol).
  • the update data on the primary volume is periodically sent to the update from the volume.
  • the primary volume logs through the log to record the location of the data that needs to be updated from the volume during the synchronization cycle. Therefore, the log can also be regarded as a location record, which records the location of the update data and the update status of each data in the location.
  • Location records can usually be recorded in a bitmap format, but are not limited thereto, and other methods are also possible. In the bitmap, "1" can be used to indicate that the data of the location has been updated, and a "0" indicates that the data of the location has not been updated.
  • the data that needs to be updated can be stored in multiple ways. It can be saved as a snapshot of the primary volume, or it can be directly saved in the log volume.
  • the remote mirror recovery method of the embodiment of the present invention compares the bits recorded by the new primary volume and the original primary volume.
  • Figure after the location that really needs data recovery, copy the update data of these locations to the original primary volume, complete the image recovery, and avoid copying all the data to the original primary volume, thus improving the image recovery speed.
  • FIG. 3 is a flowchart of a mirror recovery method according to an embodiment of the present invention.
  • Figure 3 specifically includes the steps:
  • Step Bl the primary volume sends the update data to the slave volume for update
  • the application host sends a write request to the primary volume of the primary storage device.
  • the asynchronous remote mirror program of the primary storage device writes the data of the write request to the primary volume of the primary storage device according to the location requested in the write request.
  • record in the location record for example, when the bitmap mode is adopted, the identifier of the corresponding position in the bitmap is set to "1", indicating that the data of the location has been updated. If the location of a location in the bitmap is "0", it means that the location is not updated.
  • the update data of the corresponding location is sent to the slave volume according to the information recorded in the bitmap, and is updated after being received from the volume.
  • the identifier of the corresponding location of the primary volume is set to "0" from “1".
  • the primary volume In a synchronization cycle, the primary volume only needs to synchronize the data that needs to be synchronized to the secondary volume after the update in this cycle to the secondary volume. If the primary volume is updated again during this synchronization period, the updated data will wait until the next volume. A synchronization cycle is then transferred to the slave volume, thus ensuring that the data transmitted during a synchronization cycle is consistent.
  • Step B2 The primary volume is temporarily invalidated, and the new volume is changed from the takeover service
  • the clustered software of the application host cooperates with the asynchronous remote mirroring program of the master/slave storage device, and the service of the primary volume is switched to take over from the volume, and the secondary volume becomes the new primary volume.
  • the location where the data update occurs is also recorded through the bitmap.
  • the identifier of the location record of the new primary volume is initially set to "0".
  • the identifier of the location where the data is updated is Set to "1".
  • Step B3 The original primary volume is restored, and the new primary volume obtains the bitmap of the original primary volume recorded before the failure from the original primary volume; the bitmap may also be a bitmap that sends a request to the original primary volume to obtain the original primary volume for data update.
  • Step B4 The new primary volume performs a bitwise operation on the bitmap recorded by itself and the received bitmap to obtain a location where the original primary volume has different data from the new primary volume at this time;
  • the bitmap of the record is 011
  • the data of the new primary volume from the volume is D, F
  • the bitmap of the record is 011
  • the bit or operation is performed.
  • the result is 011, which means that the data corresponding to the next two positions is different.
  • the slave volume should copy D and F to the original primary volume for updating.
  • Step B5 Copy the data of the corresponding position in the new primary volume bitmap to the original primary volume according to the operation result.
  • the data of the corresponding position in the bitmap of the new primary volume record is copied to the original primary volume, that is, the new primary volume sends the data of the corresponding position in the bitmap to the original primary volume, and the original primary volume is updated after receiving, thereby completing Image recovery.
  • the cluster software of the application host cooperates with the asynchronous remote mirroring program of the master/slave storage device, the service is switched from the new primary volume to the original primary volume.
  • FIG. 4 it is a flowchart of a method for image restoration in the second embodiment of the present invention.
  • the main difference between the second embodiment and the first embodiment is that the meanings indicated by "1" and "0" in the bitmap are different, and the corresponding subsequent comparison process uses the bitwise AND operation.
  • Figure 4 specifically includes the steps:
  • Step Cl the primary volume sends the update data to the slave volume for update
  • the application host sends a write request to the primary volume of the primary storage device.
  • the asynchronous remote mirror program of the primary storage device writes the data of the write request to the primary volume of the primary storage device according to the location requested in the write request.
  • record in the location record for example, when the bitmap mode is adopted, the identifier of the corresponding position in the bitmap is set to “0”, indicating that the data of the location has been updated. If the location of a location in the bitmap is "1", it means that the location is not updated.
  • the update data of the corresponding location is sent to the slave volume according to the information recorded in the bitmap, and is updated after being received from the volume.
  • the identifier of the corresponding location of the primary volume is set from “0" to "1".
  • the primary volume In a synchronization cycle, the primary volume only needs to synchronize the data that needs to be synchronized to the secondary volume after the update in this cycle to the secondary volume. If the primary volume is updated again during this synchronization period, the updated data will wait until the next volume. A synchronization cycle is then transferred to the slave volume, thus ensuring that the data transmitted during a synchronization cycle is consistent.
  • Step C2 the primary volume temporarily expires, and becomes a new primary volume from the volume takeover service;
  • the clustering software of the application host cooperates with the asynchronous remote mirroring program of the master/slave storage device, and the service of the primary volume is switched to take over from the volume, and the secondary volume becomes the new primary volume.
  • the location where the data update occurs is also recorded through the bitmap.
  • the identifier of the location record of the new primary volume is initially set to "1".
  • the identifier of the location where the data is updated is Set to "0".
  • Step C3 The original primary volume is restored, and the new primary volume obtains the bitmap of the original primary volume recorded before the failure from the original primary volume; the bitmap may also be a bitmap that sends a request to the original primary volume to obtain the original primary volume for data update.
  • the new primary volume after the original primary volume is restored, the new primary volume sends a request to the original primary volume, requesting to obtain a bitmap recorded before the failure in the original primary volume, and the original primary volume sends the bitmap recorded before the failure to the new primary volume.
  • Step C4 The new primary volume performs a bitwise operation on the bitmap recorded by itself and the received bitmap to obtain a location where the original primary volume has different data from the new primary volume at this time;
  • the slave volume should copy D and F to the original primary volume for updating.
  • Step C5 Copy the data of the corresponding position in the new primary volume bitmap to the original primary volume according to the operation result.
  • the data of the corresponding position in the bitmap of the new primary volume record is copied to the original primary volume, that is, the new primary volume sends the data of the corresponding position in the bitmap to the original primary volume, and the original primary volume is updated after receiving, thereby completing the image restoration.
  • the cluster software of the application host cooperates with the asynchronous remote mirroring program of the master/slave storage device, the service is switched from the new primary volume to the original primary volume.
  • the above two embodiments are to record the location of the data update and the update status of the location corresponding data, and compare the bitmap to obtain the location where the original primary volume and the new primary volume have different data, and thus only need to The data of the corresponding location is copied to the original primary volume to complete the image recovery as an example. However, it is not limited to this. It can also be recorded by other log methods, and the original primary volume and the new primary volume have different data by comparing the records of the log mode. Location, and then simply copy the data from the corresponding location to the original primary volume to complete the image recovery.
  • FIG. 5 it is a schematic structural diagram of a storage device according to an embodiment of the present invention.
  • the storage device includes: an obtaining unit 100, a recording unit 200, a comparing unit 300, and a data transmitting unit 400.
  • the obtaining unit 100 is configured to obtain a location record for another storage device to perform data update, and may periodically receive a location record sent by the another storage device for performing data update.
  • the location record records the update status of the data at each location.
  • the recording unit 200 is configured to record a location record of the data update by the storage device itself. After the storage device receives the write request from the application host, the data is updated, and the recording unit 200 records the location where the data is updated. The recording unit 200 records the location where the data is updated by the log.
  • the log can also be regarded as a location record, which records the location of the update data and the update status of each data in the location. Location records can usually be recorded in bitmap mode, but are not limited to this, and other methods are also possible.
  • a bitmap is also a sequence of binary sequences. Each bit in the sequence is "0" or "1", representing two different states.
  • the storage device adopts "1" to indicate that the data of the corresponding location has been updated, the use of "0" indicates that the data is not updated.
  • the recording unit 200 sets the identifier of the corresponding location from " ⁇ ". It is "0".
  • the comparing unit 300 is configured to compare the location record of the data update recorded by the recording unit 200 with the location record of the data update performed by another storage device acquired by the acquiring unit 100, to obtain the storage device itself and another storage device. A location with different data.
  • the data sending unit 400 is configured to send data corresponding to the location obtained by the comparing unit 300 to another storage device, and perform data update after being received by another storage device.
  • the comparison unit 300 further includes a first comparison unit 301 and a second comparison unit 302, or only one of the units.
  • the first comparison unit 301 is configured to perform bitwise OR operation on the location record of the data update recorded by the recording unit 200 and the location record of the data update performed by another storage device acquired by the acquiring unit 100, to obtain the storage device itself and Another storage device contains locations of different data, the location record is 1 indicating that the data has been updated, and 0 is indicating that the data is not updated.
  • the second comparison unit 302 is configured to perform a bitwise AND operation on the location record of the data update recorded by the recording unit 200 and the location record of the data update performed by another storage device acquired by the obtaining unit 100, to obtain the storage device itself and Another storage device contains locations of different data, the location The record is 0 to indicate that the data has been updated, and 1 to indicate that the data has not been updated.
  • the storage device further includes a requesting unit 500, configured to send a request to another storage device, requesting to acquire a location record for performing data update before another storage device fails, and receiving, by the obtaining unit 100, the pre-failure data update returned by another storage device. Location record.
  • FIG. 6 is a schematic structural diagram of a network system according to an embodiment of the present invention.
  • the network system includes a first storage device 10 and a second storage device 20, the first storage device 10 may refer to a primary storage device, and the second storage device 20 may refer to a secondary storage device.
  • the second storage device 20 has the structure shown in FIG. 5, and includes an acquisition unit 100, a recording unit 200, a comparison unit 300, a data transmission unit 400, and a request unit 500.
  • the comparison unit 300 includes a first comparison unit 301 and a second comparison unit 302. , or only include one of the units.
  • the first storage device 10 is configured to send a location record for performing data update, receive data, and update. After receiving the request sent by the second storage device 20, the first storage device 10 transmits a location record for performing data update to the second storage device 20. After the first storage device 10 receives the data sent by the second storage device 20, the data update is performed.
  • a second storage device 20 configured to acquire a location record sent by the first storage device 10 for performing data update; a location record where the second storage device 20 performs data update and a location where the first storage device 10 performs data update The record is compared to obtain a location where the second storage device 20 and the first storage device 10 contain different data; and the data corresponding to the obtained location is sent to the first storage device 10.
  • the second storage device 20 includes a first comparison unit 301 and a second comparison unit 302, or only one of the units.
  • the first comparison unit 301 is configured to perform bitwise OR operation on the location record of the data update by the second storage device 20 and the location record of the data update by the first storage device 10, to obtain the second storage device 20 and the first storage device. 10 A location containing different data, the location record is 1 indicating that the data has been updated, and 0 is indicating that the data is not updated.
  • the second comparison unit 302 is configured to perform a bitwise operation on the location record of the data update by the second storage device 20 and the location record of the data update by the first storage device 10, to obtain the second storage device 20 and the first storage device.
  • 10 Location containing different data the location record is 0 to indicate that the data has been updated, and 1 is used to indicate that the data is not updated.
  • the technical solution provided by the embodiment of the present invention is: obtaining a location record of data update by the first storage device, where the location record records an update status of data at each location; and the second storage device locates the local end Recording is compared with the acquired location record of the first storage device to obtain a location containing different data from the first storage device; and copying the data corresponding to the location to the first storage device.
  • the second storage device and the first storage device are obtained by adding a process of comparing the location record of the data update of the second storage device with the location record of the data update performed by the first storage device.
  • the location with different data, the data corresponding to these locations is the data that needs to be copied to the first storage device to complete the image recovery, thus overcoming the defect that the entire data needs to be copied in the prior art, thereby improving the mirror recovery speed.
  • Asynchronous remote mirroring can also be applied to environments with high disaster tolerance levels, and the data to be copied is reduced, the burden on the link is reduced, and the fault tolerance of the system is improved.
  • a bit OR operation may be used, or a bit and operation may be used, if in the bitmap Use 1 to indicate that the data has been updated, and 0 to indicate that the data is not updated. The bit or operation is used. If 0 is used in the bitmap to indicate that the data has been updated, and 1 is used to indicate that the data is not updated, the bitwise AND operation is used.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Quality & Reliability (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Description

镜像恢复方法、 存储设备及网络系统
本申请要求于 2007 年 6 月 7 日提交中国专利局、 申请号为 200710108636.8, 发明名称为"镜像恢复方法、 存储设备及网络系统 "的中国专 利申请的优先权, 其全部内容通过引用结合在本申请中。
技术领域
本发明涉及计算机网络技术领域,具体涉及一种镜像恢复方法、存储设备 及网络系统。
背景技术
目前容灾备份技术愈加重要。远程镜像是建立容灾备份系统的其中一种技 术。镜像是在两个或多个存储设备上产生同一个数据的镜像视图的信息存储过 程。 远程镜像可分为同步远程镜像和异步远程镜像。
同步远程镜像 (同步复制技术),是指主站的应用主机每发出一个写请求, 必须等待数据完全写入主存储设备的主卷和从存储设备的从卷之后,才能继续 执行。 同步远程镜像一般适用于通信链路较短的情况(通常几十公里以内)。 异步远程镜像(异步复制技术), 是主站的应用主机连续发送多个写请求时, 只需等待数据写入主存储设备的主卷之后即可继续执行,此后再由主存储设备 的主卷与从存储设备的从卷进行连续的多次数据同步,即两个存储设备之间允 许滞后多个操作。异步远程镜像适用于通信链路较长的情况(通常为几百甚至 上千里)。
异步远程镜像过程涉及到镜像恢复问题。在异步的远程镜像中, 主卷上的 更新数据是周期性的发送到从卷上进行更新的。 更新过程中, 主卷暂时失效, 从卷成为新主卷, 在原主卷恢复后, 将新主卷的数据完全复制到原主卷, 完成 镜像恢复。
请参阅图 1 , 是现有技术异步远程镜像恢复流程图。
现有异步远程镜像的恢复技术,是在主卷失效之后,从卷接管主卷的业务, 成为新主卷。 经过一段时间之后, 主卷得到恢复, 业务需要从新主卷切换回原 主卷。 此时, 将新主卷的数据完全复制到原主卷上去, 然后将业务切换回原主 卷。
图 1中具体包括步骤: 步骤 Al、 主卷将更新数据发送到从卷进行更新;
应用主机向主存储设备的主卷发送写请求,主存储设备的异步远程镜像程 序收到写请求后,根据写请求中要求写入的位置将写请求的更新数据写入主存 储设备的主卷中, 并将位图中对应的位置的标识置为 "1", 表示该位置的数据 已经更新, 如果位图中的某位置的标识为" 0", 则表示该位置的数据没有更新。 位图也就是指一串二进制序列, 序列上的每一位为" 0"或" 1", 分别表示两种不 同的状态。
主卷完成数据更新后,在同步周期到来时,按照位图中记录的信息将对应 位置的更新数据发送至从卷, 由从卷接收后进行更新。
步骤 A2、 主卷暂时失效, 从卷接管业务成为新主卷;
当主卷暂时失效时 , 经过应用主机的集群软件与主 /从存储设备的异步远 程镜像程序的配合, 主卷的业务切换至从卷进行接管, 从卷成为新主卷。
从卷成为新主卷后 , 同样通过位图来记录发生数据更新的位置。
步骤 A3、 原主卷恢复, 将新主卷的数据全部复制到原主卷。
从存储设备定时检测主存储设备的原主卷是否正常,如果检测到原主卷正 常, 则认为原主卷已经恢复。 原主卷恢复后, 将新主卷的数据全部复制到原主 卷上去, 即新主卷将全部数据发送到原主卷, 原主卷接收后进行更新, 从而完 成镜像恢复。 然后, 经过应用主机的集群软件与主 /从存储设备的异步远程镜 像程序的配合, 业务再由新主卷切换到原主卷。
发明人在实现本发明过程中,发现现有的镜像恢复过程中,镜像恢复速度 比较慢, 花费时间比较长, 不能满足容灾级别要求高的应用。
发明内容
本发明实施例要解决的技术问题是现有的镜像恢复方法、存储设备及网络 系统中镜像恢复速度比较慢的问题。
为解决上述技术问题, 本发明实施例是通过以下技术方案实现: 本发明实施例提供一种镜像恢复方法, 包括: 获取第一存储设备进行数据 更新的位置记录, 所述位置记录记录各位置上数据的更新状态; 第二存储设备 将本端的位置记录与所述获取的第一存储设备的位置记录进行比较,获得与第 一存储设备含有不同数据的位置; 将所述位置对应的数据复制到第一存储设 备。
本发明实施例提供一种存储设备, 包括: 获取单元, 用于获取另一存储设 备进行数据更新的位置记录, 所述位置记录记录各位置上数据的更新状态; 记 录单元, 用于记录所述存储设备本端的位置记录; 比较单元, 用于将所述记录 单元记录的本端的位置记录与所述获取单元获取的所述另一存储设备的位置 记录进行比较,获得与所述另一存储设备含有不同数据的位置;数据发送单元, 用于将所述比较单元获得的所述位置对应的数据发送给所述另一存储设备。
本发明实施例提供一种网络系统, 包括: 第一存储设备, 用于发送进行数 据更新的位置记录,接收数据并进行更新, 所述位置记录记录各位置上数据的 更新状态; 第二存储设备, 用于获取所述第一存储设备发送的位置记录, 将本 端的位置记录与所述获取的第一存储设备的位置记录进行比较,获得与所述第 一存储设备含有不同数据的位置, 将所述位置对应的数据发送到第一存储设 备。
以上技术方案可以看出,本发明实施例提供的技术方案是: 获取第一存储 设备进行数据更新的位置记录, 所述位置记录记录各位置上数据的更新状态; 第二存储设备将本端的位置记录与所述获取的第一存储设备的位置记录进行 比较,获得与第一存储设备含有不同数据的位置; 将所述位置对应的数据复制 到第一存储设备。 因为本发明实施例方案中,通过增加对第二存储设备进行数 据更新的位置记录与所述第一存储设备进行数据更新的位置记录进行比较的 过程,得出第二存储设备与第一存储设备含有不同数据的位置,这些位置对应 的数据也就是真正需要复制到第一存储设备完成镜像恢复的数据,这样就不需 要象现有技术那样将全部数据进行复制,从而提高了镜像恢复速度,也使异步 远程镜像能应用于容灾级别高的环境, 并且需复制的数据减少了,也减轻了链 路的负担, 提高了系统的容错性。
附图说明
图 1是现有技术异步远程镜像恢复流程图;
图 2是本发明实施例镜像恢复方法的一个应用环境示意图;
图 3是本发明实施例一镜像恢复方法流程图;
图 4是本发明实施例二镜像恢复方法流程图; 图 5是本发明实施例存储设备的结构示意图;
图 6是本发明实施例网络系统的结构示意图。
具体实施方式
现有的镜像恢复方法的镜像恢复速度比较慢,本发明实施例提供一种镜像 恢复方法, 能够提高镜像恢复速度。
请参阅图 2, 是本发明实施例镜像恢复方法的一个应用环境示意图。
图 2 中描述了异步远程镜像的其中一种组网情况, 其中包括两台应用主 机, 它们通过广域网 WAN ( Wide Area Network )或局域网 LAN ( Local Area Network ) 互联, 在两台应用主机上安装集群软件而构成一个高可用测试 HA ( High Availability )的环境。 应用主机上运行的业务可以是数据库服务、 邮件 服务以及文件服务等企业级应用。组网中还包括两台存储设备, 即主存储设备 和从存储设备。应用主机的业务数据保存在主存储设备的主卷和从存储设备的 从卷中。 应用主机通过存储区域网络 SAN ( Storage Area Network )与主、 从 存储设备互联,承载的协议可以是 FC协议(Fibre Channel, 光纤通道)、 iSCSI ( internet SCSI, 互联网 SCSI技术)协议和 SAS协议( Serial Attached SCSI, 串行 SCSI技术)等。 SCSI是 Small Computer System Interface的英文简写 , 中 文一般称为小型计算机系统接口。应用主机的业务数据先写入到主存储设备的 主卷中, 然后由主卷再将数据镜像至从存储设备的从卷。主存储设备和从存储 设备的数据链路可以是 FC ( Fibre Channel, 光纤通道)、 DWDM ( Dense Wave length Division Multiplexing, 密集波分复用 )、 ATM ( Asynchronous Transfer Mode, 异步传输模式)或 IP ( Internet Protocol, 互联网协议)等传输线路。
异步的远程镜像中,主卷上的更新数据是周期性的发送到从卷上进行更新 的。 主卷通过日志来记录同步周期内需要向从卷更新的数据的位置。 因此, 日 志也可以看成是位置记录,该位置记录记录了更新数据的位置及位置上各数据 的更新状态等。位置记录通常可以采用位图方式进行记录但不局限于此,也可 以采用其他方式。 位图中可以用 "1"表示该位置的数据已更新, 用" 0"表示该位 置的数据未更新。在一个同步周期内,需要更新的数据的存放方式可以有多种, 可以是以主卷快照的方式进行保存, 也可以是直接将数据保存在日志卷中。
本发明实施例的远程镜像恢复方法是通过比较新主卷和原主卷记录的位 图,得出真正需要进行数据恢复的位置后,将这些位置的更新数据复制到原主 卷上, 完成镜像恢复, 避免了需将全部数据复制到原主卷上, 从而提高了镜像 恢复速度。
请参阅图 3, 是本发明实施例一镜像恢复方法流程图。
图 3中具体包括步骤:
步骤 Bl、 主卷将更新数据发送到从卷进行更新;
应用主机向主存储设备的主卷发送写请求,主存储设备的异步远程镜像程 序收到写请求后,根据写请求中要求写入的位置将写请求的数据写入主存储设 备的主卷中, 并在位置记录中进行记录, 例如在采用位图方式时将位图中对应 的位置的标识置为" 1", 表示该位置的数据已经更新。 如果位图中的某位置的 标识为 "0", 则表示该位置没有更新。
主卷完成数据更新后,在同步周期到来时,按照位图中记录的信息将对应 位置的更新数据发送至从卷, 由从卷接收后进行更新。 当主卷将对应位置的更 新数据发送至从卷后 , 主卷对应位置的标识从 "1"置为" 0"。
在一个同步周期内,主卷只需要将本周期内更新后需要同步到从卷的数据 发送给从卷即可, 如果此同步周期内, 主卷再发生更新, 那么这些更新的数据 将等到下一个同步周期再传送到从卷,这样就保证了一个同步周期内传送的数 据是一致的。
步骤 B2、 主卷暂时失效, 从卷接管业务成为新主卷;
当主卷暂时失效时, 经过应用主机的集群软件与主 /从存储设备的异步远 程镜像程序的配合, 主卷的业务切换至从卷进行接管, 从卷成为新主卷。
从卷成为新主卷后, 同样通过位图来记录发生数据更新的位置。新主卷的 位置记录的标识初始都是置为" 0", 当新主卷接受应用主机向新主卷发送写请 求后, 新主卷的数据发生更新时, 数据被更新的位置的标识被置为 "1"。
步骤 B3、原主卷恢复, 新主卷从原主卷获取原主卷在失效前记录的位图; 位图 , 也可以是向原主卷发出请求以获取原主卷进行数据更新的位图。
本实施例中是原主卷恢复后,新主卷向原主卷发送请求,请求获取原主卷 中在失效前记录的位图 , 原主卷将失效前记录的位图发送给新主卷。 步骤 B4、 新主卷将自身记录的位图和接收的位图进行位或运算, 得到原 主卷此时与新主卷有不同数据的位置;
举个例子, 假如原主卷的数据是 A、 B、 C, 记录的位图是 011, 而从卷即 新主卷的数据是 、 D、 F, 记录的位图是 011, 那么进行位或运算后的结果就 是 011, 表示后面两个位置对应的数据是不同的, 从卷应该将 D、 F都复制到 原主卷进行更新。
步骤 B5、根据运算结果, 将新主卷位图中对应位置的数据复制到原主卷。 才艮据运算结果,将新主卷记录的位图中对应位置的数据复制到原主卷, 即 新主卷将位图中对应位置的数据发送到原主卷,原主卷接收后进行更新,从而 完成镜像恢复。 然后, 经过应用主机的集群软件与主 /从存储设备的异步远程 镜像程序的配合, 业务再由新主卷切换到原主卷。
请参阅图 4, 是本发明实施例二镜像恢复方法流程图。 实施例二和实施例 一的主要区别是位图中 "1"和" 0"表示的含义不同,相应的后面的比较过程采用 位与运算。
图 4中具体包括步骤:
步骤 Cl、 主卷将更新数据发送到从卷进行更新;
应用主机向主存储设备的主卷发送写请求,主存储设备的异步远程镜像程 序收到写请求后,根据写请求中要求写入的位置将写请求的数据写入主存储设 备的主卷中, 并在位置记录中进行记录, 例如在采用位图方式时将位图中对应 的位置的标识置为" 0", 表示该位置的数据已经更新。 如果位图中的某位置的 标识为 "1", 则表示该位置没有更新。
主卷完成数据更新后,在同步周期到来时,按照位图中记录的信息将对应 位置的更新数据发送至从卷, 由从卷接收后进行更新。 当主卷将对应位置的更 新数据发送至从卷后 , 主卷对应位置的标识从" 0"置为" 1"。
在一个同步周期内,主卷只需要将本周期内更新后需要同步到从卷的数据 发送给从卷即可, 如果此同步周期内, 主卷再发生更新, 那么这些更新的数据 将等到下一个同步周期再传送到从卷,这样就保证了一个同步周期内传送的数 据是一致的。
步骤 C2、 主卷暂时失效, 从卷接管业务成为新主卷; 当主卷暂时失效时 , 经过应用主机的集群软件与主 /从存储设备的异步远 程镜像程序的配合, 主卷的业务切换至从卷进行接管, 从卷成为新主卷。
从卷成为新主卷后, 同样通过位图来记录发生数据更新的位置。新主卷的 位置记录的标识初始都是置为" 1", 当新主卷接受应用主机向新主卷发送写请 求后, 新主卷的数据发生更新时, 数据被更新的位置的标识被置为" 0"。
步骤 C3、原主卷恢复, 新主卷从原主卷获取原主卷在失效前记录的位图; 位图 , 也可以是向原主卷发出请求以获取原主卷进行数据更新的位图。
本实施例中是原主卷恢复后,新主卷向原主卷发送请求,请求获取原主卷 中在失效前记录的位图, 原主卷将失效前记录的位图发送给新主卷。
步骤 C4、 新主卷将自身记录的位图和接收的位图进行位与运算, 得到原 主卷此时与新主卷有不同数据的位置;
举个例子, 假如原主卷的数据是 A、 B、 C, 记录的位图是 011, 而从卷即 新主卷的数据是0、 F、 C, 记录的位图是 001, 那么进行位与运算后的结果就 是 001, 表示前面两个位置对应的数据是不同的, 从卷应该将 D、 F都复制到 原主卷进行更新。
步骤 C5、根据运算结果, 将新主卷位图中对应位置的数据复制到原主卷。 根据运算结果,将新主卷记录的位图中对应位置的数据复制到原主卷, 即 新主卷将位图中对应位置的数据发送到原主卷,原主卷接收后进行更新,从而 完成镜像恢复。 然后, 经过应用主机的集群软件与主 /从存储设备的异步远程 镜像程序的配合, 业务再由新主卷切换到原主卷。
需要说明的是,上述两个实施例是以位图记录发生数据更新的位置及位置 对应数据的更新状态, 并通过比较位图获得原主卷与新主卷有不同数据的位 置,进而只需将对应位置的数据复制到原主卷完成镜像恢复为例进行说明但不 局限于此,也可以用其他日志方式进行记录, 并通过比较该日志方式的记录来 获得原主卷与新主卷有不同数据的位置,进而只需将对应位置的数据复制到原 主卷完成镜像恢复。
上述内容伴细介绍了本发明实施例的镜像恢复方法,相应的,本发明实施 例提供一种存储设备和网络系统。 请参阅图 5 , 是本发明实施例存储设备的结构示意图。
如图 5所示, 该存储设备包括: 获取单元 100、 记录单元 200、 比较单元 300和数据发送单元 400。
获取单元 100, 用于获取另一存储设备进行数据更新的位置记录, 可以是 定期接收所述另一存储设备发送的进行数据更新的位置记录。所述位置记录记 录各位置上数据的更新状态。
记录单元 200, 用于记录存储设备自身进行数据更新的位置记录。 存储设 备接收到应用主机的写请求后,进行数据更新, 并由记录单元 200记录进行数 据更新的位置。记录单元 200通过日志来记录进行数据更新的位置。 日志也可 以看成是位置记录,该位置记录记录了更新数据的位置及位置上各数据的更新 状态等。位置记录通常可以采用位图方式进行记录但不局限于此,也可以采用 其他方式。 位图也就是指一串二进制序列, 序列上的每一位为" 0"或" 1", 分别 表示两种不同的状态。 假如存储设备采用 "1"表示对应位置的数据已更新, 采 用" 0"表示未更新, 当存储设备将更新数据发送至另一存储设备后, 记录单元 200将对应位置的标识由 "Γ,置为 "0"。
比较单元 300, 用于将所述记录单元 200记录的进行数据更新的位置记录 与所述获取单元 100获取的另一存储设备进行数据更新的位置记录进行比较, 获得存储设备自身与另一存储设备含有不同数据的位置。
数据发送单元 400, 用于将所述比较单元 300获得的位置所对应的数据发 送给另一存储设备, 由另一存储设备接收后进行数据更新。
比较单元 300进一步包括第一比较单元 301和第二比较单元 302, 或只包 括其中一个单元。
第一比较单元 301 , 用于将所述记录单元 200记录的进行数据更新的位置 记录与所述获取单元 100获取的另一存储设备进行数据更新的位置记录进行 位或运算,获得存储设备自身与另一存储设备含有不同数据的位置, 所述位置 记录是采用 1表示数据已更新, 采用 0表示数据未更新。
第二比较单元 302, 用于将所述记录单元 200记录的进行数据更新的位置 记录与所述获取单元 100获取的另一存储设备进行数据更新的位置记录进行 位与运算,获得存储设备自身与另一存储设备含有不同数据的位置, 所述位置 记录是采用 0表示数据已更新, 采用 1表示数据未更新。
存储设备进一步包括请求单元 500, 用于向另一存储设备发送请求, 请求 获取另一存储设备失效前进行数据更新的位置记录,并由获取单元 100接收另 一存储设备返回的失效前进行数据更新的位置记录。
请参阅图 6, 是本发明实施例网络系统的结构示意图。
该网络系统包括第一存储设备 10和第二存储设备 20, 第一存储设备 10 可以是指主存储设备, 第二存储设备 20可以是指从存储设备。 第二存储设备 20具有图 5所示的结构, 包括获取单元 100、 记录单元 200、 比较单元 300、 数据发送单元 400和请求单元 500, 比较单元 300包括第一比较单元 301和第 二比较单元 302, 或只包括其中一个单元。
第一存储设备 10, 用于发送进行数据更新的位置记录, 接收数据并进行 更新。 第一存储设备 10接收第二存储设备 20发送的请求后, 向第二存储设备 20发送进行数据更新的位置记录。 第一存储设备 10接收第二存储设备 20发 送的数据后, 进行数据更新。
第二存储设备 20, 用于获取第一存储设备 10发送的进行数据更新的位置 记录; 将所述第二存储设备 20进行数据更新的位置记录与所述第一存储设备 10进行数据更新的位置记录进行比较, 获得所述第二存储设备 20与第一存储 设备 10含有不同数据的位置; 将获得的位置所对应的数据发送到第一存储设 备 10。
所述第二存储设备 20包括第一比较单元 301和第二比较单元 302, 或只 包括其中一个单元。
第一比较单元 301, 用于将第二存储设备 20进行数据更新的位置记录与 所述第一存储设备 10进行数据更新的位置记录进行位或运算, 获得第二存储 设备 20与第一存储设备 10含有不同数据的位置,所述位置记录是采用 1表示 数据已更新 , 采用 0表示数据未更新。
第二比较单元 302, 用于将第二存储设备 20进行数据更新的位置记录与 所述第一存储设备 10进行数据更新的位置记录进行位与运算, 获得第二存储 设备 20与第一存储设备 10含有不同数据的位置,所述位置记录是采用 0表示 数据已更新, 采用 1表示数据未更新。 以上技术方案可以看出,本发明实施例提供的技术方案是: 获取第一存储 设备进行数据更新的位置记录, 所述位置记录记录各位置上数据的更新状态; 第二存储设备将本端的位置记录与所述获取的第一存储设备的位置记录进行 比较,获得与第一存储设备含有不同数据的位置; 将所述位置对应的数据复制 到第一存储设备。 因为本发明实施例方案中,通过增加对第二存储设备进行数 据更新的位置记录与所述第一存储设备进行数据更新的位置记录进行比较的 过程,得出第二存储设备与第一存储设备含有不同数据的位置,这些位置对应 的数据也就是真正需要复制到第一存储设备完成镜像恢复的数据,这样就克服 了现有技术中需将全部数据进行复制的缺陷,从而提高了镜像恢复速度,也使 异步远程镜像能应用于容灾级别高的环境, 并且需复制的数据减少了,也减轻 了链路的负担, 提高了系统的容错性。
进一步的,在将第二存储设备进行数据更新的位置记录与所述第一存储设 备进行数据更新的位置记录进行比较过程中,可以采用位或运算,也可以采用 位与运算, 如果位图中采用 1表示数据已更新, 用 0表示数据未更新, 则采用 的是位或运算, 如果位图中采用 0表示数据已更新, 用 1表示数据未更新, 则 采用的是位与运算。
以上对本发明实施例所提供的镜像恢复方法、存储设备及网络系统进行了 述, 以上实施例的说明只是用于帮助理解本发明实施例的方法及其核心思想; 同时, 对于本领域的一般技术人员, 依据本发明实施例的思想, 在具体实施方 式及应用范围上均会有改变之处, 综上所述,本说明书内容不应理解为对本发 明的限制。

Claims

权 利 要 求
1、 一种镜像恢复方法, 其特征在于, 包括:
获取第一存储设备进行数据更新的位置记录,所述位置记录记录各位置上 数据的更新状态;
第二存储设备将本端的位置记录与所述获取的第一存储设备的位置记录 进行比较, 获得与第一存储设备含有不同数据的位置;
将所述位置对应的数据复制到第一存储设备。
2、 根据权利要求 1所述的镜像恢复方法, 其特征在于:
所述第二存储设备将本端的位置记录与所述获取的第一存储设备的位置 记录进行比较具体为:
所述第二存储设备将本端的位置记录与所述获取的第一存储设备的位置 记录进行位或运算, 所述位置记录是采用 1 表示对应位置的数据已更新的状 态, 采用 0表示对应位置的数据未更新的状态。
3、 根据权利要求 1所述的镜像恢复方法, 其特征在于:
所述第二存储设备将本端的位置记录与所述获取的第一存储设备的位置 记录进行比较具体为:
所述第二存储设备将本端的位置记录与所述获取的第一存储设备的位置 记录进行位与运算, 所述位置记录是采用 0表示对应位置的数据已更新的状 态, 采用 1表示对应位置的数据未更新的状态。
4、 根据权利要求 1、 2或 3所述的镜像恢复方法, 其特征在于: 所述获取第一存储设备进行数据更新的位置记录具体为:
所述第二存储设备向所述第一存储设备发送获取所述第一存储设备失效 前的位置 i己录的请求,
接收所述第一存储设备返回的失效前的位置记录; 或者,
所述第二存储设备定期接收所述第一存储设备发送的失效前的位置记录。
5、 根据权利要求 1所述的镜像恢复方法, 其特征在于:
所述第二存储设备的位置记录与第一存储设备的位置记录采用位图进行 记录。
6、 一种存储设备, 其特征在于, 包括:
获取单元, 用于获取另一存储设备进行数据更新的位置记录, 所述位置记 录记录各位置上数据的更新状态;
记录单元, 用于记录所述存储设备本端的位置记录;
比较单元,用于将所述记录单元记录的本端的位置记录与所述获取单元获 取的所述另一存储设备的位置记录进行比较,获得与所述另一存储设备含有不 同数据的位置;
数据发送单元,用于将所述比较单元获得的所述位置对应的数据发送给所 述另一存储设备。
7、 根据权利要求 6所述的存储设备, 其特征在于:
所述比较单元包括第一比较单元,用于将所述记录单元记录的本端的位置 记录与所述获取单元获取的所述另一存储设备的位置记录进行位或运算,获得 与所述另一存储设备含有不同数据的位置,所述位置记录是采用 1表示对应位 置的数据已更新的状态, 采用 0表示对应位置的数据未更新的状态。
8、 根据权利要求 6或 7所述的存储设备, 其特征在于:
所述比较单元包括第二比较单元,用于将所述记录单元记录的本端的位置 记录与所述获取单元获取的所述另一存储设备的位置记录进行位与运算,获得 与所述另一存储设备含有不同数据的位置,所述位置记录是采用 0表示对应位 置的数据已更新的状态, 采用 1表示对应位置的数据未更新的状态。
9、 根据权利要求 6或 7所述的存储设备, 其特征在于, 所述存储设备还 包括:
请求单元, 用于请求获取所述另一存储设备失效前的位置记录, 并由所述 获取单元接收所述另一存储设备返回的失效前进行数据更新的位置记录。
10、 一种网络系统, 其特征在于, 包括:
第一存储设备,用于发送进行数据更新的位置记录,接收数据并进行更新, 所述位置记录记录各位置上数据的更新状态;
第二存储设备, 用于获取所述第一存储设备发送的位置记录,将本端的位 置记录与所述获取的第一存储设备的位置记录进行比较,获得与所述第一存储 设备含有不同数据的位置 , 将所述位置对应的数据发送到第一存储设备。
11、 根据权利要求 10所述的网络系统, 其特征在于:
所述第二存储设备包括第一比较单元,用于将本端的位置记录与所述第一 存储设备的位置记录进行位或运算, 获得与第一存储设备含有不同数据的位 置, 所述位置记录是采用 1表示对应位置的数据已更新的状态, 采用 0表示对 应位置的数据未更新的状态。
12、 根据权利要求 10或 11所述的网络系统, 其特征在于:
所述第二存储设备包括第二比较单元,用于将本端的位置记录与所述第一 存储设备的位置记录进行位与运算, 获得与第一存储设备含有不同数据的位 置, 所述位置记录是采用 0表示对应位置的数据已更新的状态, 采用 1表示对 应位置的数据未更新的状态。
13、 根据权利要求 10或 11所述的网络系统, 其特征在于:
所述第二存储设备还包括请求单元,用于请求获取所述第一存储设备失效 前的位置记录,所述第二存储设备接收所述另一存储设备根据所述请求返回的 失效前进行数据更新的位置记录。
PCT/CN2008/070879 2007-06-07 2008-05-05 Procédé de récupération d'image, équipement de stockage et système réseau WO2008148319A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CNB2007101086368A CN100454265C (zh) 2007-06-07 2007-06-07 镜像恢复方法、存储设备及网络系统
CN200710108636.8 2007-06-07

Publications (1)

Publication Number Publication Date
WO2008148319A1 true WO2008148319A1 (fr) 2008-12-11

Family

ID=38880358

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2008/070879 WO2008148319A1 (fr) 2007-06-07 2008-05-05 Procédé de récupération d'image, équipement de stockage et système réseau

Country Status (2)

Country Link
CN (1) CN100454265C (zh)
WO (1) WO2008148319A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103064763A (zh) * 2012-12-27 2013-04-24 华为技术有限公司 一种数据备份方法及相关设备、系统
CN109828950A (zh) * 2019-01-29 2019-05-31 深圳前海达闼云端智能科技有限公司 一种数据保护方法、管理系统、服务器及可读存储介质

Families Citing this family (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN100454265C (zh) * 2007-06-07 2009-01-21 华为技术有限公司 镜像恢复方法、存储设备及网络系统
CN101808137B (zh) * 2010-03-29 2014-09-03 成都市华为赛门铁克科技有限公司 数据传输方法、装置和系统
CN102306115B (zh) * 2011-05-20 2014-01-08 华为数字技术(成都)有限公司 异步远程复制方法、系统及设备
CN102662800A (zh) * 2012-04-20 2012-09-12 浪潮电子信息产业股份有限公司 一种保证远程异步镜像业务不中断的方法
EP3062226B1 (en) 2013-12-12 2018-02-21 Huawei Technologies Co., Ltd. Data replication method and storage system
CN105988901B (zh) * 2013-12-12 2019-06-18 华为技术有限公司 数据复制方法及存储系统
CN105208104B (zh) * 2015-08-26 2018-10-02 浪潮(北京)电子信息产业有限公司 一种远程镜像数据一致性检测方法及装置
US10416905B2 (en) * 2017-02-09 2019-09-17 Hewlett Packard Enterprise Development Lp Modifying membership of replication groups via journal operations
CN107038091B (zh) * 2017-03-29 2019-07-26 国网山东省电力公司信息通信公司 一种基于异步远程镜像的数据安全性保护系统与电力应用系统数据保护方法
CN107329701A (zh) * 2017-06-29 2017-11-07 郑州云海信息技术有限公司 一种存储系统中双活卷的创建方法、装置及系统
CN110990191B (zh) * 2019-11-01 2022-06-17 武汉船舶通信研究所(中国船舶重工集团公司第七二二研究所) 一种基于镜像存储的数据恢复方法及系统
CN113010259A (zh) * 2019-12-20 2021-06-22 上海云轴信息科技有限公司 两节点超融合一体机的存储方法及设备

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1057533A (zh) * 1990-06-21 1992-01-01 国际商业机器公司 恢复受奇偶校验保护数据的方法和装置
CN1454349A (zh) * 2000-06-07 2003-11-05 处理存储器有限公司 主存储器事务处理系统中的高度并行日志和恢复操作的方法和系统
JP2004145854A (ja) * 2002-06-11 2004-05-20 Hitachi Ltd データバックアップと回復のための方法と装置
CN1742262A (zh) * 2003-01-31 2006-03-01 松下电器产业株式会社 用于产生差分数据的装置和方法、用于恢复更新后数据的装置和方法、及其程序
CN101067796A (zh) * 2007-06-07 2007-11-07 华为技术有限公司 镜像恢复方法、存储设备及网络系统

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1057533A (zh) * 1990-06-21 1992-01-01 国际商业机器公司 恢复受奇偶校验保护数据的方法和装置
CN1454349A (zh) * 2000-06-07 2003-11-05 处理存储器有限公司 主存储器事务处理系统中的高度并行日志和恢复操作的方法和系统
JP2004145854A (ja) * 2002-06-11 2004-05-20 Hitachi Ltd データバックアップと回復のための方法と装置
CN1742262A (zh) * 2003-01-31 2006-03-01 松下电器产业株式会社 用于产生差分数据的装置和方法、用于恢复更新后数据的装置和方法、及其程序
CN101067796A (zh) * 2007-06-07 2007-11-07 华为技术有限公司 镜像恢复方法、存储设备及网络系统

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103064763A (zh) * 2012-12-27 2013-04-24 华为技术有限公司 一种数据备份方法及相关设备、系统
CN109828950A (zh) * 2019-01-29 2019-05-31 深圳前海达闼云端智能科技有限公司 一种数据保护方法、管理系统、服务器及可读存储介质

Also Published As

Publication number Publication date
CN101067796A (zh) 2007-11-07
CN100454265C (zh) 2009-01-21

Similar Documents

Publication Publication Date Title
WO2008148319A1 (fr) Procédé de récupération d'image, équipement de stockage et système réseau
US11567674B2 (en) Low overhead resynchronization snapshot creation and utilization
TWI307035B (en) Method and system for backing up remote mirror data on internet
CN108496153B (zh) 同步复制关系的非破坏性基线和再同步
US9934242B2 (en) Replication of data between mirrored data sites
JP5559821B2 (ja) データを記憶する方法、データをミラーリングする方法、命令シーケンスを担持する機械読み取り可能な媒体、および、当該方法をコンピュータに実行させるためのプログラム
CN103226502B (zh) 一种数据灾备控制系统及数据恢复方法
WO2016070375A1 (zh) 一种分布式存储复制系统和方法
US9740572B1 (en) Replication of xcopy command
US7421549B2 (en) Method and apparatus of remote copy for multiple storage subsystems
US8935498B1 (en) Splitter based hot migration
US7600087B2 (en) Distributed remote copy system
JP4473612B2 (ja) リモートコピーシステム
JP3718471B2 (ja) 完全な再ミラーリングのないクラッシュリカバリ
US7565572B2 (en) Method for rolling back from snapshot with log
US8285824B2 (en) Storage system and data replication method that refuses one or more requests for changing the first logical configuration information until the first storage apparatus and second storage apparatus are synchronized
JP4461147B2 (ja) リモートデータミラーリングを用いたクラスタデータベース
US7831550B1 (en) Propagating results of a volume-changing operation to replicated nodes
JP2014532921A (ja) 高可用性クラスタにおけるスプリット・ブレイン耐性フェイルオーバ
JP2007052772A (ja) データ複製システム並びにデータ移行方法
US7107355B2 (en) High availability lightweight directory access protocol service
JPWO2013046352A1 (ja) 計算機システム、データ管理方法及びデータ管理プログラム
US20230409227A1 (en) Resilient implementation of client file operations and replication
US8117493B1 (en) Fast recovery in data mirroring techniques
CN113905054B (zh) 基于RDMA的Kudu集群数据同步方法、装置、系统

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: 08734235

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 08734235

Country of ref document: EP

Kind code of ref document: A1