TW201403315A - Cloud storage system and method thereof - Google Patents

Abstract

The present invention provides a cloud storage system and a method thereof. The cloud storage system is composed of a client end, a network storage device, and an intermediate data server. The intermediate data server stores pointers in a storage bank thereof. The network storage device stores files and stores pointers in a storage bank thereof. A request from the client end is directly transferred to a primary network storage device. The primary network storage device determines one of the intermediate data server, the primary network storage device itself, and other affiliated network storage devices thereof to perform subsequent processing according to the contents of the request. The system and method of the present invention can increase file access performance, dynamically adjust allocation of storage resources, and prevent the intermediate data server from becoming a bottleneck of performance.

Description

Cloud storage system and method thereof

The invention relates to a file storage system stored in a cloud storage system through a network, and can effectively improve file access performance when a client performs file operations. In addition, the system can dynamically adjust the file and indicator configuration according to the state of the storage resource, so the intermediary data server can be prevented from becoming a performance bottleneck, especially a cloud storage system and method thereof through the network.

Many storage devices are placed inside a computer system, such as a hard disk, to allow clients to store data inside the computer system. Other storage devices can also be connected to the system, such as floppy disks, CDs or mobile disks, to allow customers. The storage operation is performed outside the system. By using an external storage device, the client can conveniently carry digital content to be stored. However, due to the limited number of hard disks in the computer system, the storage capacity cannot be expanded arbitrarily. However, the mobile disk is easy to carry, but because of its small size, it is easy to be lost and neglected to be carried, so the mobile disk is inconvenient to use. .

Therefore, the use of the cloud storage system for storage operations overcomes these difficulties, especially in today's rapid increase in network transmission rate, only need to use the network to connect to the cloud storage system, you can easily and quickly access a variety of files. operating. The cloud storage system has many advantages such as cost saving, simple deployment, convenient maintenance, fast switching of the system according to actual needs, and mutual interference when the virtual disks are used. However, the current cloud storage system does not have a complete system and method, which can effectively improve file access performance, dynamically configure storage resources, and prevent the intermediary data server from becoming a performance bottleneck. In view of the above, the present invention is directed to the above problems, and proposes a cloud storage system and method improvement thereof to effectively solve the conventional problems.

The main purpose of the present invention is to provide a cloud storage system and a method thereof, which use a network connection to access files in a cloud storage system, and improve storage access performance in a cloud storage system to satisfy a client storage service. Product Quality requirements.

Another object of the present invention is to provide a cloud storage system and method thereof, which can dynamically adjust file and indicator configurations according to the state of the cloud storage system and can improve storage resources by distributing a plurality of indicators to the network storage device. Use efficiency and reduce mediation data server load.

To achieve the above objective, the present invention provides a cloud storage system in which a primary network storage device is responsible for accepting a request from a client and connecting an attached network storage device to control an attached network storage device. Both the primary network storage device and the attached network storage device can store files and store the indicators in their storage. The primary network storage device can directly process the client's request or forward the request to the affiliate network storage device or the intermediary data server for subsequent processing. An intermediary data server manages the indicators and stores them in the repository, and can transmit the related information of the network storage device where the client files are located to the client, so that the client can directly access the files.

In addition, the present invention also provides a cloud storage method for improving file access performance and avoiding mediation by dynamically performing load balancing of storage resources, file transfer, file copy and indicator moving, and distributing indicators to network storage devices. The data server becomes a performance bottleneck. The details, technical contents, features, and effects achieved by the present invention will become more apparent from the detailed description of the embodiments.

The invention provides a cloud storage system and a method thereof to improve file access performance to meet the storage service quality requirements of the client. Dynamically perform load balancing, file movement, and file copying by detecting system status to ensure efficient file access. In addition, the overloading mechanism of the mediation data server is avoided by the indicator moving mechanism and the method of distributing the indicators to the network storage device. The system can also provide a storage capacity with no upper limit for the client by increasing the number of network storage devices. In addition, file access operations via the network can improve convenience and efficiency.

The first figure is the system architecture diagram of the present invention, as shown in the figure, the cloud storage The system 10 includes an intermediary data server 12 that includes a repository 20 that is responsible for storing the indicators 22. Indicator 22 contains the directory name, file name, and file-related mediation information, such as attributes such as owner, access rights, setup time, number of accesses per unit time, and file destination. A client 14 is responsible for sending file access requests and then performing actual file access. The primary network storage device 16 directly accepts access requests from the client 14, which can be used to store files and have a repository 20 for storing the metrics 22. The affiliate network storage device 18 is managed by the primary network storage device 16 for storing files and storing the indicators 22 in the repository 20. The mediation data server 12, the primary network storage device 16, and the attached network storage device 18 are each configured with a dedicated repository 20 for storing the metrics 22, so that the metrics 22 can be stored arbitrarily to reduce the load on the mediation data server 12. Finally, two affiliated network storage devices 18 are exemplified in this embodiment.

Please refer to the first figure and cooperate with the second figure. The second figure is a flow chart of the file uploading step of the present invention. As shown in step S10, the file upload request is sent by the client 14 with the directory name and file name in the primary network storage device 16. Then, proceeding to step S12, all the affiliated network storage devices 18 can notify the primary network storage device 16 of the latest state when the storage resource changes exceed a certain allowable value, and the primary network storage device 16 can be based on The information reported by each of the attached network storage devices 18, for example, the available storage space, the network bandwidth usage, and the disk input/output read/write frequency, dynamically adjusts each of the attached network storages in the storage resource rotation configuration table. The device 18 is best suited to store capacity quotas, thereby balancing the load of the entire cloud storage system 10. Next, proceeding to step S14, the primary network storage device 16 determines the storage location of the file, that is, the preset affiliate network storage device 18, based on the directory name and file name accompanying the request. The simple approach is to uniformly map all of the attached network storage devices 18 in the system with the directory name. For example, there are currently 36 auxiliary network storage devices 18 in the cloud storage system 10, and the file has a directory name at the beginning of a, and its default affiliate network storage device 18 is set as the first affiliate network storage device 18. So on and so forth. Next, if the file has a beginning of 0 The subsidiary network storage device 18 whose name is preset is set as the 27th affiliate network storage device 18. The identifier of the attached network storage device 18 can be implemented using an IP address or by a unique naming scheme by the primary network storage device 16. Next, the primary network storage device 16 compares whether the secondary network storage device 18 in the rotation is the same as the secondary network storage device 18 to which the request is preset. If not, the process proceeds to step S16, and the primary network storage device 16 forwards the request along with the preset affiliate network storage device 18's network connection information to the secondary network storage device 18 in the round. The network connection information is the type of internet protocol, the type of transmission control protocol, the communication number, the encryption type, and the account password to establish an internet connection. Then, proceeding to step S18, the satellite network storage device 18 is responsible for transmitting the response message to the client 14, and then the client 14 can directly associate with the network storage device 18 in the round according to the network connection information carried in the response message. Establish a network connection. Next, proceeding to step S20, the client 14 can upload the file to the affiliated network storage device 18 in the rotation. After the file transfer is completed, the process proceeds to step S22, in which the attached network storage device 18 can transmit the file index 22 to the preset affiliate network according to the network connection information provided by the previous primary network storage device 16. The storage device 18, in addition to the common attributes of the file, in particular, the file destination attribute will be designated as the network connection information of the attached network storage device 18 in the rotation instead of the preset auxiliary network storage. Device 18 itself. That is to say, in this case, the files and indicators 22 are stored in different network storage devices. Finally, in step S24, the preset affiliate network storage device 18 updates the indicator 22 to its repository 20. On the other hand, if the comparison result is the same, the process proceeds to step S26, and the primary network storage device 16 forwards the request to the secondary network storage device 18 in the round. Then, proceeding to step S28, the satellite network storage device 18 transmits a response message to the client 14 in the round, and then the client 14 can establish a network connection according to the information provided by the response message. Next, proceeding to step S30, the client 14 directly uploads the file to the attached network storage device 18 in the rotation. After the file is received, the process proceeds to step S32, in which the satellite network storage device 18 updates the file index 22 to its own repository 20. In this case, both the file and the indicator 22 are stored in Same network storage device. In general, the file and indicator 22 are stored in the same network storage device, and the access performance is better because there is no need to forward the request.

Please refer to the first figure and cooperate with the third figure. The third figure is a flowchart of the file downloading step of the present invention. As shown in step S34, the file download request is sent by the client 14 with the directory name and file name in the primary network storage device 16. The primary network storage device 16 first queries the repository 20 and finds that the file cannot be found. Corresponding indicator 22, proceeding to step S36, forwarding the request to the intermediary data server 12 and the preset affiliate network storage device 18, because the index 22 corresponding to the file may be moved from the preset affiliate network storage device 18. The intermediary data server 12 is arranged to prevent the preset affiliate network storage device 18 from storing too many indicators. The next two will query their exclusive repository 20 separately. Since there is only one network location for the index 22 corresponding to the file, if any of the queries is successful, the process proceeds to step S38, and the mediation data server 12 or the preset The affiliate network storage device 18 directly transmits the response message to the client 14, and then the client 14 can directly establish a network with the attached network storage device 18 storing the file according to the network connection information carried by the response message. Connect the line and download the file. Next, proceeding to step S40, both the intermediary data server 12 and the preset affiliate network storage device 18 return the query result to the primary network storage device 16. Then, proceeding to step S42, the primary network storage device 16 compares the two query results. If any of the query results is successful, the process proceeds to step S44, and after the client 14 completes the file download, the auxiliary network of the downloaded file is stored. The storage device 18 checks whether there is a corresponding indicator 22 for the file in its repository 20. If so, the file access time is updated to the corresponding indicator 22 in its repository 20. On the contrary, information such as the directory name, file name, and file access time is transmitted to the mediation data server 12 and the preset affiliate network storage device 18. If the corresponding indicator 22 is found to be stored in its repository 20, the corresponding indicator 22 in the repository 20 is updated, otherwise no processing is performed. If both of the queries fail, the process proceeds to step S46, indicating that the file name may be incorrect or the file does not exist, and the primary network storage device 16 is responsible for returning an error message to the client 14, and then displaying the file download operation. failure.

Please refer to the first figure and the fourth figure. The fourth figure is a flow chart of the load balancing step of the present invention. As shown in step S48, initially, each of the accessory network storage devices 18 periodically or event-triggered to notify the user of the storage resource state change to the primary network storage device 16, and then the primary network storage device 16 will be based on all The storage resource status provided by the adjunct network storage device 18 determines the storage capacity quota for each of the adjunct network storage devices 18. Next, proceeding to step S50, the primary network storage device 16 establishes a storage resource rotation configuration table according to the obtained information, and the table specifies the storage capacity of each of the subsidiary network storage devices 18 at each round, Information such as identifiers and order of rotation. Then, proceeding to step S52, once the storage capacity of the attached network storage device 18 has reached the storage setting preset value, the primary network storage device 16 is actively sent. Finally, proceeding to step S54, the primary network storage device 16 will find the next affiliate network storage device 18 to replace the secondary network storage in the round according to the storage resource rotation configuration table after receiving the notification from the attached network storage device 18. Device 18. Once the storage resource rotation configuration table has been executed, the primary network storage device 16 re-plans the storage resource rotation configuration table used in the next round according to the storage resource update information received during the rotation, so that the cloud storage system can reflect the storage in time. Resource changes and the ability to load balance.

Please refer to the first figure and cooperate with the fifth figure. The fifth figure is a flow chart of the file moving step of the present invention. As shown in step S56, the affiliate network storage device 18 finds that its own storage space has increased due to factors such as file deletion, file transfer, or storage capacity expansion. Next, proceeding to step S58, the affiliate network storage device 18 evaluates and finds the appropriate indicator 22 for file regression, and can take the index with higher file access operation priority as the priority, so that the file access speed can be accelerated. Of course, the index 22 of the file regression means that the corresponding file does not actually exist in the affiliate network storage device 18 itself, but only stores its index 22. In other words, file access cannot be directly performed, but subsequent processing must be performed by forwarding the access request to the attached network storage device 18 where the file is actually stored. Undoubtedly, file movement can improve file access performance. Once selected for the appropriate indicator 22 Then, proceeding to step S60, the file transfer operation is started. Of course, before the file is moved, the affiliate network storage device 18 that originally stored the file and the attached network storage device 18 of the file transfer destination can coordinate and complete the file transfer planning table before the file is moved to avoid affecting the existing storage service. quality. Once the file transfer is completed, the process proceeds to step S62, and the attached network storage device 18 of the file transfer destination modifies the file destination information in the corresponding indicator 22 in the repository 20, and changes from the previously stored affiliate network storage device 18 to Yourself. Since the storage resource status of the attached network storage device 18 involved in the file migration has different degrees of impact, the process proceeds to step S64, and it is determined according to the preset condition whether the storage resource status needs to be updated to the primary network storage device 16, To plan a storage resource rotation configuration table that meets the status quo.

Please refer to the first figure and cooperate with the sixth figure. The sixth figure is a flow chart of the steps for establishing the file copy of the present invention. First, as shown in step S66, the affiliate network storage device 18 detects the file exceeding the hotspot threshold. The hotspot threshold is defined herein as the upper limit of the number of file accesses per unit time. Next, proceeding to step S68, if the file and the corresponding indicator 22 are stored in the same affiliate network storage device 18, the attached network storage device 18 first changes the file destination attribute in the indicator 22 to be transmitted to itself. The network connection information, as for the indicator 22 originally in the repository, remains unchanged, and then the file and indicator 22 are transmitted to the primary network storage device 16. On the contrary, the attached network storage device 18 storing the indicator 22 directly transmits the indicator 22 to the attached network storage device 18 where the file is located, and then the attached network storage device 18 is responsible for transmitting the file and receiving index 22 To the primary network storage device 16. Finally, proceeding to step S70, the main network storage device 16 stores the received file and adds itself to the file destination attribute in the corresponding indicator 22, which means that the file is stored in two different locations, that is, there are two copies. . Finally, the indicator 22 is updated to the repository 20. Once the client 14 wants to access the file, the primary network storage device 16 can decide to process it by itself or by another affiliated network storage device 18 storing the file to improve the file storage according to the file access situation. Take performance and balance the load. If the file access frequency is low to some extent, the primary network storage device 16 can It is determined whether to delete the file and the corresponding indicator 22 according to the state of its own storage resource. Once the delete operation is decided, the primary network storage device 16 deletes the archive and the corresponding indicator 22 directly without additional processing.

Please refer to the first figure and the seventh figure. The seventh figure is a flow chart of the indicator configuration adjustment steps of the present invention. First, as shown in step S72, the number of indicators 22 stored in an affiliate network storage device 18 exceeds a preset value, which may affect the file access performance, because the number of comparisons at the time of the query may be increased. Next, proceeding to step S74, the affiliate network storage device 18 selects suitable indicators for moving until the number of indicators 22 is lower than a preset value. If the indicator 22 does not exist in the same affiliate network storage device 18 as the file, the indicator 22 will be listed as a priority transfer object. Next, proceeding to step S76, the affiliate network storage device 18 transmits the indicator 22 for determining the move to the mediation data server 12. Finally, proceeding to step S78, after the indicator 22 is successfully moved, the affiliate network storage device 18 removes the relevant indicator 22 from the repository 20, and the intermediary data server 12 adds the received indicator 22 to its repository 20. If the number of indicators 22 on the mediation data server 12 is too large, the single media data server 12 load can be dispersed by constructing the mediation data server 12 cluster or the index 22 can be moved to the subsidiary network whose number of indicators 22 is lower than the preset value. On the road storage device 18.

The above is only the preferred embodiment of the present invention and is not intended to limit the scope of the present invention. Therefore, any changes or modifications of the features and spirits of the present invention should be included in the scope of the present invention.

10‧‧‧Cloud Storage System

12‧‧‧Intermediary data server

14‧‧‧ Client

16‧‧‧Main network storage device

18‧‧‧Affiliated network storage device

20‧‧‧Repository

22‧‧‧ indicators

The first figure is a schematic diagram of the system architecture of the present invention.

The second figure is a flow chart of the file uploading step of the present invention.

The third figure is a flow chart of the file downloading steps of the present invention.

The fourth figure is a flow chart of the load balancing step of the present invention.

The fifth figure is a flow chart of the file moving step of the present invention.

The sixth figure is a flow chart of the steps for establishing a file copy of the present invention.

The seventh figure is a flow chart of the adjustment step of the indicator configuration of the present invention.

10‧‧‧Cloud Storage System

12‧‧‧Intermediary data server

14‧‧‧ Client

16‧‧‧Main network storage device

18‧‧‧Affiliated network storage device

20‧‧‧Repository

22‧‧‧ indicators

Claims (13)

A cloud storage system includes: a primary network storage device to receive a request from at least one client; and an intermediary data server to query a file storage location; and further comprising at least one affiliate network storage device to receive the primary network Control of the road storage device. The cloud storage system of claim 1, wherein the primary network storage device, the secondary network storage device, and the intermediary data server each include a repository for storing indicators. The cloud storage system of claim 1, wherein the request is an archive upload instruction and an archive download instruction. As requested in item 2, the indicator includes attributes such as directory name, file name, owner, file access time, and file destination. A cloud storage method includes: a primary network storage device to receive a request from at least one client; and an intermediary data server to query a file storage location; and at least one auxiliary network storage device to receive the primary network Control of the road storage device. The cloud storage method of claim 5, wherein the request is an archive upload instruction and an archive download instruction. The cloud storage method of claim 6, wherein the file uploading instruction first finds the secondary network storage device in the rotation and compares it with the preset affiliate network storage device. If they are the same, the request is forwarded and the response message is transmitted by the attached network storage device in the rotation, and then the file is uploaded and the corresponding indicator in the repository is updated. Conversely, the preset affiliate network storage device network connection information is transmitted, and then the response message is transmitted and the file upload is performed by the attached network storage device in the rotation. Finally, the indicator is transmitted to the preset affiliate network storage device to update the repository. The cloud storage method of claim 6, wherein the file downloading instruction first queries the in-resource index of the main network storage device to confirm whether the file exists, and if so, directly downloads the file. Otherwise, the request is forwarded to the mediation data server and the default affiliate network storage device. Once the file location is confirmed, Send a response message directly. The next two will transfer the query results to the main network storage device. If the query fails, an error message is sent, otherwise the indicator will be updated to the corresponding repository after the file is downloaded. The cloud storage method as claimed in claim 5, comprising: load balancing, file moving, file copy creation, and indicator configuration adjustment. The cloud storage method of claim 9, wherein the primary network storage device determines the storage capacity quota of the configuration according to the storage resource status of the attached network storage device and establishes a storage resource rotation configuration table accordingly. Once the secondary network storage device in the rotation reaches the storage setting default value, a notification is issued and the new affiliate network storage device is found for replacement. The cloud storage method of claim 9, wherein the accessory network storage device finds that the storage resource has been improved, and then selects a suitable indicator for file regression, and then performs file migration. Then, the corresponding indicator in the repository is modified and the state of the storage resource is updated to construct a storage resource rotation configuration table. The cloud storage method of claim 9, wherein the accessory network storage device detects the file exceeding the hotspot threshold, and transmits the file and adjusts the file destination attribute of the corresponding indicator to the primary network storage device, thereby The indicator is updated to the repository. The cloud storage method of claim 9, when the number of indicators owned by the attached network storage device exceeds a preset value, the appropriate indicator is selected for moving, and then the indicator is transmitted to the intermediary data server for update storage. Library.