CN103559319A - Cache synchronization method and equipment for distributed cluster file system - Google Patents

Abstract

Embodiments of the present invention provide a cache synchronization method and device for a distributed cluster file system. By defining a global cache table for the target file, the global cache table contains the modification records of each node to the target file. When the first node updates the target file When the file is written and accessed, the latest global cache table is obtained according to the information of the second node that has recently modified the target file in the write lock, and the second node is obtained according to the latest modification record of the target file by the second node in the latest global cache table. The modification data of the target file by the two nodes, when the two nodes write different data of the target file interactively at the same time, the data in their respective caches are valid, and subsequent reads can be obtained directly from the cache, when the two nodes write the same data interactively When the modified data needs to be obtained from another node, the transfer of data between networks is much more efficient than writing and reading disks once, so the cluster file system is greatly improved compared with the existing methods.

Description

Cache synchronization method and device for distributed cluster file system

technical field

Embodiments of the present invention relate to data communication technology, and in particular to a cache synchronization method and device for a distributed cluster file system.

Background technique

The distributed cluster file system can solve the problem of massive data storage. In the cluster, multiple devices are combined to provide external services, and multiple nodes (servers) are used to replace one node to complete the task. This not only improves the processing capacity, but also ensures high reliability. The symmetrically distributed cluster file system has a globally unified namespace. The storage space is visible and identical to each node. Metadata and data can be distributed to any node in the cluster. Of course, the system can also be accessed from any node. All data stored.

Current storage systems are divided into at least two levels of storage: storage physical media and caches created from main memory. In order to improve the access performance of the distributed cluster file system, there will be data caches in each node. When the client reads and accesses some data, the node receiving the request first checks whether the data is stored in the cache of the node. If so , return directly, if not, send an I/O request to the disk, read the data from the disk into the cache and save it temporarily, in case the next access does not need to read from the disk, read directly from the cache, to Improve read performance. Similarly, when a client writes and accesses some data, the node receiving the request first writes the data into the cache instead of writing it directly to the disk. Data is written to disk to improve write performance. When multiple nodes in the system interact to read and write access to the same file, how to efficiently ensure data consistency is a key problem that the system needs to solve.

In the prior art, for a distributed cluster file system, distributed locks must be used to ensure data consistency. Before accessing data, it is first necessary to apply for a distributed lock corresponding to the data. Distributed locks include read locks and write locks. Read locks can be concurrent, and write locks are exclusive. That is, multiple nodes can read the same part of data concurrently. However, if a node is modifying this part of data, other nodes must wait until the modification is completed. Access, the accessed data includes the latest modified data. A specific example will be used to illustrate the following. Figure 1 is a schematic diagram of the file reading and writing process in the prior art. As shown in Figure 1, node 1 first modifies the 0th data block of file 1, because no node has applied for it before. The write lock corresponding to file 1 is in the initialization state, and the lock management device is directly authorized to node 1. After obtaining the write lock, node 1 modifies the 0th data block, and the modified data is stored in the cache. After node 2 writes access to file 1, it needs to modify the fourth data block of file 1. Distributed lock finds that node 1 has applied for a write lock, so it reclaims the write lock of file 1 from node 1, and node 1 receives the command to reclaim the write lock After that, the cached data of file 1 is first downloaded to the disk. Since node 2 applies for a write lock, it means that node 2 wants to modify the data of file 1, so the cached data related to file 1 on node 1 may not be the latest , it needs to be invalidated, and then return the lock authority to the lock management device. The lock management device authorizes the lock authority to node 2. After node 2 obtains the write lock, it modifies the corresponding data to the cache. After the modification, node 1 reads and accesses file 1 The 0th data block of , as before, the lock management device reclaims the write lock from node 2. During the recovery process, the data of node 2 is written to the disk. Since node 1 applies for a read lock, reading will not modify the data, so the cache of node 2 It is still valid, that is, there is no need to invalidate the corresponding cache, and then the lock management device authorizes the read lock to node 1, and node 1 finds the corresponding data from the cache, but finds that the data is invalid, indicating that the data may not be up-to-date. disk read.

From the above example, we can see that by writing the latest data to the disk to ensure the consistency of the data accessed by each node, there will be many disadvantages as follows: If the nodes in the cluster read and write the same file interactively, this will cause File data is frequently read and written to the disk, and the cache does not have any acceleration effect, but increases the complexity of the system, and frequent disk reading and writing will increase the delay of the system, which will greatly affect performance.

Contents of the invention

Embodiments of the present invention provide a cache synchronization method and device for a distributed cluster file system, which can avoid increased system delay caused by frequent disk reading and writing, and improve system performance.

The first invention of the present invention provides a cache synchronization method for a distributed cluster file system, including:

When the client performs write access to the target file on the first node, the first node applies to the lock management device for the write lock corresponding to the target file, and the write lock includes the latest modification of the target file The information of the second node;

The first node obtains the latest global cache table corresponding to the target file according to the information of the second node, and the latest modification record of the target file by the second node is stored in the latest global cache table;

The first node modifies the target file according to the latest modification record of the second node to the target file, saves the modified data in the cache of the first node, and stores the first The modification record of the target file by the node is updated to the latest global cache table.

In a first possible implementation manner of the first aspect of the present invention, the first node modifies the target file according to the latest modification record of the second node to the target file, including:

The first node determines whether to acquire the modification data of the target file by the second node according to the latest modification record of the target file by the second node;

If so, the first node obtains the modification data of the target file by the second node, and stores the obtained modification data of the target file in the cache of the first node, so that the first node The data of the target file stored on the node and the second node are consistent, the data is read from the cache of the first node, and the target file is modified;

If not, the first node reads data from the cache of the first node, and modifies the target file.

In a second possible implementation manner of the first aspect of the present invention, the target file includes at least two data blocks, and the latest modification record of the target file by the second node includes: the second node pair Information about the latest modified data block of the target file;

The first node determines whether to acquire the modification data of the target file by the second node according to the latest modification record of the second node to the target file, including:

The first node determines that the data block that the first node performs write access to the target file is identical to the second node's Whether the latest modified data block of the target file is the same data block;

If yes, the first node obtains the modification data of the target file by the second node.

In a third possible implementation manner of the first aspect of the present invention, the first node obtains the latest global cache table according to the information of the second node, including:

The first node sends a global cache table synchronization request message to the second node;

The first node receives the global cache table synchronization response message sent by the second node, the global cache table synchronization response message includes the latest global cache table, and the latest global cache table stores the second The latest modification record of the target file by the node.

In a fourth possible implementation manner of the first aspect of the present invention, the first node obtains the modification data of the target file by the second node according to the latest modification record of the target file by the second node ,include:

The first node sends a data synchronization request message to the second node according to the latest modification record of the target file by the second node in the latest global cache table;

The first node receives the data synchronization response message sent by the second node, and the data synchronization response message includes the modification data of the target file by the second node.

In a fifth possible implementation manner of the first aspect of the present invention, the first node performs a write operation on the target file, saves the modified data in the cache of the first node, and stores the After the first node updates the modification record of the target file to the latest global cache table, it also includes:

When the first node performs read access to the target file, it applies to the lock management device for a read lock corresponding to the target file, and the read lock includes the latest modification of the target file. information of the first node;

The first node reads the target file from the cache of the first node.

In a sixth possible implementation manner of the first aspect of the present invention, the first node performs a write operation on the target file, saves the modified data in the cache of the first node, and stores the After a node updates the modification record of the target file to the latest global cache table, it also includes:

The first node receives the global cache table synchronization request message sent by the second node, and the global cache table synchronization request message is sent to the second node when the second node needs to perform read or write access to the target file. sent by the first node;

The first node returns a global cache table synchronization response message to the second node according to the global cache table synchronization request message, and the global cache table synchronization response message includes the updated latest global cache table.

In a seventh possible implementation manner of the first aspect of the present invention, the first node returns a global cache table synchronization response message to the second node or the third node according to the global cache table synchronization request message After that, also include:

The first node receives the data synchronization request sent by the second node, and the data synchronization request message is that the second node updates the target file by the first node according to the updated latest global cache table. sent by the modification record;

The first node sends a data synchronization response to the second node, and the data synchronization response includes the modification data of the target file by the first node, so that the second node converts the first node The modified data of the target file is stored in the cache of the second node, and the data can be read from the cache of the second node when the second node performs read access or write access to the data. Target file.

In an eighth possible implementation manner of the first aspect of the present invention, the first node applies to a lock management device for a write lock corresponding to the target file, including:

The first node sends a write lock application request corresponding to the target file to the lock management device, so that the lock management device reclaims the write lock to the second node;

The first node receives the write lock authorization response sent by the lock management device.

The second aspect of the present invention provides a cache synchronization method of a distributed cluster target file system, including:

The lock management device receives the write lock application request corresponding to the target file sent by the first node, and the write lock application request is that when the client performs write access to the target file, the first node sends to the lock management device sent;

The lock management device determines that the second node has applied for the write lock corresponding to the target file, and the target file has been newly modified, then the lock management device sends the write lock corresponding to the target file to the second node. A lock reclaim command, reclaiming the write lock corresponding to the target file from the second node, and the lock management device does not reclaim the write lock corresponding to the target file from the second node The data of the target file in the cache is written to disk and invalidated;

The lock management apparatus sends a write lock authorization response to the first node, and the write lock includes information of the second node.

In a first possible implementation manner of the second aspect of the present invention, the lock management device sends a write lock authorization response to the first node, and after the write lock includes the information of the second node, it also includes :

The lock management device receives a read lock application request corresponding to the target file sent by the first node;

The lock management device determines that the first node has applied for the write lock corresponding to the target file, and the target file has been newly modified, then the lock management device sends the target file corresponding to the first node to the first node. The write lock recovery command of the first node recovers the write lock corresponding to the target file from the first node. When the lock management device recovers the write lock corresponding to the target file from the first node, it does not The data of the target file in the cache of a node is written to disk and invalidated;

The lock management device sends a read lock authorization response to the first node, and the read lock includes information of the first node, so that the first node reads the Describe the target file.

In a second possible implementation manner of the second aspect of the present invention, the lock management device sends a write lock authorization response to the first node, and after the write lock includes the information of the second node, it also includes :

The lock management device receives a read lock or write lock application request corresponding to the target file sent by the second node;

The lock management device determines that the first node has applied for the write lock corresponding to the target file, and the target file has been newly modified, then the lock management device sends the target file corresponding to the first node to the first node. The write lock recovery command of the first node recovers the write lock corresponding to the target file from the first node. When the lock management device recovers the write lock corresponding to the target file from the first node, it does not The data of the target file in the cache of a node is written to disk and invalidated;

The lock management device sends a read lock or write lock authorization response to the second node, and the read lock or write lock includes the information of the first node, so that the second node according to the first node information to get the latest global cache table.

The third aspect of the present invention provides a cache synchronization method of a distributed cluster target file system, including:

The second node receives the global cache table synchronization request message sent by the first node;

The second node returns a global cache table synchronization response message to the first node according to the global cache table synchronization request message, the global cache table synchronization response message includes the latest global cache table, and the latest global The latest modification record of the target file by the second node is saved in the cache table.

In the first possible implementation manner of the third aspect of the present invention, after the second node returns a global cache table synchronization response message to the first node according to the global cache table synchronization request message, it further includes:

The second node receives the data synchronization request message sent by the first node, and the data synchronization request message is the latest update of the target file by the second node in the latest global cache table by the first node. Amended records sent;

The second node returns a data synchronization response message to the first node according to the data synchronization request message, and the data synchronization response message includes modification data of the target file by the second node.

A fourth aspect of the present invention provides a node, where the node is a first node, including:

A lock application module, configured to apply to the lock management device for a write lock corresponding to the target file when the client performs write access to the target file on the first node, and the write lock includes the update of the target file. Modified information of the second node;

A cache table acquisition module, configured to acquire the latest global cache table corresponding to the target file according to the information of the second node, where the latest modification record of the second node to the target file is stored in the latest global cache table ;

A file modification module, configured to modify the target file according to the latest modification record of the second node to the target file, save the modified data in the cache of the first node, and save the modified data in the cache of the first node A node updates the modification record of the target file to the latest global cache table.

In a first possible implementation manner of the fourth aspect of the present invention, the file modification module includes a judgment unit, a data acquisition unit, and a file modification unit;

The judging unit is configured to determine whether to acquire modification data of the target file by the second node according to the latest modification record of the target file by the second node;

If so, the obtaining unit obtains the modification data of the target file by the second node, and stores the obtained modification data of the target file in the cache of the first node, so that the first node Keeping consistent with the data of the target file stored on the second node, the file modification unit reads the data from the cache of the first node, and modifies the target file;

If not, the file modifying unit reads data from the cache of the first node, and modifies the target file.

In a second possible implementation manner of the fourth aspect of the present invention, the target file includes at least two data blocks, and the latest modification record of the target file by the second node includes: the second node pair Information about the latest modified data block of the target file;

The judging unit is specifically used for:

According to the information of the data block that the second node has recently modified the target file, determine that the data block that the first node performs write access to the target file is the same as the data block that the second node performs write access to the target file Whether the latest modified data block is the same data block;

If yes, the data acquisition unit acquires modification data of the target file by the second node.

In a third possible implementation manner of the fourth aspect of the present invention, the cache table obtaining module is specifically configured to:

sending a global cache table synchronization request message to the second node;

Receive a global cache table synchronization response message sent by the second node, the global cache table synchronization response message includes the latest global cache table, and the latest global cache table stores the A record of the last modification of the file.

In a fourth possible implementation manner of the fourth aspect of the present invention, the data acquisition unit is specifically configured to:

Sending a data synchronization request message to the second node according to the latest modification record of the target file by the second node in the latest global cache table;

Receive a data synchronization response message sent by the second node, where the data synchronization response message includes modification data of the target file by the second node.

In a fifth possible implementation manner of the fourth aspect of the present invention, it further includes: a file reading module;

The lock application module is further configured to: when the first node performs read access to the target file, apply to the lock management device for a read lock corresponding to the target file, and the read lock includes the Information about the first node whose target file has been newly modified;

The file reading module is configured to read the target file from the cache of the first node.

In a sixth possible implementation manner of the fourth aspect of the present invention, it also includes:

The receiving module is configured to receive the global cache table synchronization request message sent by the second node, the global cache table synchronization request message is sent to the second node when the second node needs to perform read access or write access to the target file sent by the first node;

A sending module, configured to return a global cache table synchronization response message to the second node according to the global cache table synchronization request message, where the global cache table synchronization response message includes the updated latest global cache table.

In a seventh possible implementation manner of the fourth aspect of the present invention, the receiving module is further configured to:

receiving a data synchronization request sent by the second node, where the data synchronization request message is sent by the second node according to the modification record of the target file by the first node in the updated latest global cache table ;

The sending module is further configured to: send a data synchronization response to the second node, the data synchronization response includes the modification data of the target file by the first node, so that the second node will The modified data of the target file by the first node is stored in the cache of the second node, and can be read from the cache of the second node when the second node performs read access or write access to the data Get the target file.

In an eighth possible implementation manner of the fourth aspect of the present invention, the lock application module is specifically configured to:

sending a write lock application request corresponding to the target file to the lock management device, so that the lock management device reclaims the write lock to the second node;

Receive the write lock authorization response sent by the lock management device.

A fifth aspect of the present invention is a lock management device, comprising:

The receiving module is configured to receive the write lock application request corresponding to the target file sent by the first node, the write lock application request is when the client performs write access to the target file, the first node sends to the lock sent by the management device;

A determining module, configured to determine that the second node has applied for a write lock corresponding to the target file, and that the second node has made the latest modification to the target file;

a sending module, configured to send a write lock recovery command corresponding to the target file to the second node, and recover the write lock corresponding to the target file from the second node; When the node reclaims the write lock corresponding to the target file, it does not write and invalidate the data of the target file in the cache of the second node;

The sending module is further configured to: send a write lock authorization response to the first node, where the write lock includes information of the second node.

In a first possible implementation manner of the fifth aspect of the present invention, the receiving module is further configured to:

receiving a read lock application request corresponding to the target file sent by the first node;

The determining module is further configured to: determine that the first node has applied for a write lock corresponding to the target file, and the first node has made the latest modification to the target file;

The sending module is further configured to: send a write lock recovery command corresponding to the target file to the first node, and recover the write lock corresponding to the target file from the first node, and the lock management device sends the command to the first node. When the first node reclaims the write lock corresponding to the target file, it does not write and invalidate the data of the target file in the cache of the first node;

The sending module is further configured to: send a read lock authorization response to the first node, where the read lock includes information of the first node, so that the first node retrieves the first node from the cache of the first node Read the object file.

In a second possible implementation manner of the fifth aspect of the present invention, the receiving module is further configured to:

receiving a read lock or write lock application request corresponding to the target file sent by the second node;

The determining module is further configured to: determine that the first node has applied for a write lock corresponding to the target file, and the first node has made the latest modification to the target file;

The sending module is further configured to: send a write lock recovery command corresponding to the target file to the first node, and recover the write lock corresponding to the target file from the first node, and the lock management device sends the command to the first node. When the first node reclaims the write lock corresponding to the target file, it does not write and invalidate the data of the target file in the cache of the first node;

The sending module is further configured to: send a read lock or write lock authorization response to the second node, and the read lock or write lock includes the information of the first node, so that the second node according to the The information of the first node is obtained from the latest global cache table.

A sixth aspect of the present invention provides a node, where the node is a second node, including:

a receiving module, configured to receive the global cache table synchronization request message sent by the first node;

A sending module, configured to return a global cache table synchronization response message to the first node according to the global cache table synchronization request message, where the global cache table synchronization response message includes the latest global cache table, and the latest global cache table The latest modification record of the target file by the second node is saved in the cache table.

In a first possible implementation manner of the sixth aspect of the present invention, the receiving module is further configured to:

receiving a data synchronization request message sent by the first node, where the data synchronization request message is sent by the first node according to the latest modification record of the target file by the second node in the latest global cache table;

The sending module is further configured to: return a data synchronization response message to the first node according to the data synchronization request message, and the data synchronization response message includes modification data of the target file by the second node.

Embodiments of the present invention provide a cache synchronization method and device for a distributed cluster file system. By defining a global cache table for the target file, the global cache table contains the modification records of each node to the target file. When the first node updates the target file When the file is written and accessed, the latest global cache table is obtained according to the information of the second node that has carried out the latest modification to the target file in the write lock, and the second node is obtained according to the latest modification record of the target file by the second node in the latest global cache table. The modification data of the target file by the two nodes, when two nodes write different data of the target file interactively at the same time, the data in the respective caches are valid, and subsequent reads can be obtained directly from the cache, when the two nodes write interactively In the case of the same data, the modified data needs to be obtained from another node. The transmission of data between networks is much more efficient than writing and reading disks once. Therefore, whether the same data or different data is cross-modified, the performance of the cluster file system is better than that of the current one. There are major improvements in both ways.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description These are some embodiments of the present invention. For those skilled in the art, other drawings can also be obtained according to these drawings without any creative effort.

Fig. 1 is a schematic diagram of the process of reading and writing files in the prior art;

Figure 2 is a schematic diagram of the physical architecture of the distributed cluster file system;

3 is a flow chart of Embodiment 1 of the cache synchronization method of the distributed cluster file system of the present invention;

4 is a flow chart of Embodiment 2 of the cache synchronization method of the distributed cluster file system of the present invention;

5 is a flow chart of Embodiment 3 of the cache synchronization method of the distributed cluster file system of the present invention;

6 is a flow chart of Embodiment 4 of the cache synchronization method of the distributed cluster file system of the present invention;

Fig. 7 is a schematic structural diagram of the first node of the present invention;

FIG. 8 is a schematic structural diagram of Embodiment 2 of the first node of the present invention;

9 is a schematic structural diagram of Embodiment 1 of the lock management device of the present invention;

FIG. 10 is a schematic structural diagram of Embodiment 1 of the second node of the present invention;

Fig. 11 is a schematic structural diagram of Embodiment 3 of the first node of the present invention;

Fig. 12 is a schematic structural diagram of Embodiment 2 of the lock management device of the present invention;

FIG. 13 is a schematic structural diagram of Embodiment 2 of the second node of the present invention.

Detailed ways

In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments It is a part of embodiments of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

Before introducing the various embodiments of the present invention, first briefly introduce the distributed cluster file system. Figure 2 is a schematic diagram of the physical architecture of the distributed cluster file system. As shown in Figure 2, the distributed cluster file system has six servers, respectively Node 1, Node 2, Node 3, Storage Node 1, Storage Node 2, Storage Node 3, Node 1, Node 2, and Node 3 are mainly used to process data read requests and write requests sent by clients, and Storage Node 1, Storage node 2 and storage node 3 are mainly used to store data. The data stored by node 1, node 2 and node 3 can all be stored on storage node 1, storage node 2 and storage node 3. The distributed cluster shown in Figure 2 The physical structure of the file system is only used as an example. In the actual application process, there may be more nodes, and the methods provided by the embodiments of the present invention are applicable to all distributed cluster file systems and are not affected by the distributed cluster file system. The impact of the hardware structure of the file system.

The distributed cluster file system has a globally unified namespace. The storage space is visible and the same to each node. Metadata and data can be distributed to any node in the cluster. Of course, any node can also access the system stored in the system. For all data, take Figure 2 as an example, storage node 1, storage node 2, and storage node 3 together constitute a storage space, and nodes 1, 2, and 3 can access all data on these three storage nodes. For distributed cluster file systems, distributed locks must be used to ensure data consistency. Distributed locks are hereinafter referred to as locks. Before accessing data, nodes first need to apply to the lock management device for the lock of the file corresponding to the data. The lock management device is used to manage the corresponding locks of all files in the system, and is responsible for allocating locks to each node and reclaiming locks to each node. Each file has a read lock and a write lock. In fact, the read lock and write lock of a file can be regarded as a lock. When the lock management device authorizes the lock to the node, it sets a level in the lock, and the level of the write lock It is higher than the level of read lock. If the node applies for a write lock, set the lock level to the level corresponding to the write lock. The lock also includes the object authorized by the lock. If the lock is to be authorized to the first node, the lock management device Write the information of the first node in the authorization object in the lock. Among them, the read lock can be concurrent, and the write lock is exclusive, that is, multiple nodes can read the same part of the data concurrently, but if a node is modifying the part of the data, other nodes must wait until the modification is completed before they can access the part of the data.

FIG. 3 is a flow chart of Embodiment 1 of the cache synchronization method for a distributed cluster file system according to the present invention. As shown in FIG. 3 , the method provided in this embodiment may include the following steps:

Step 101. When the client performs write access to the target file on the first node, the first node applies to the lock management device for a write lock corresponding to the target file.

When the client performs write access to the target file on the first node, the first node first applies to the lock management device for the write lock corresponding to the target file. In this embodiment, when the lock management device authorizes the write lock to the first node , the write lock includes the information of the second node that has made the latest modification to the target file.

Wherein, the first node applies to the lock management device for a write lock corresponding to the target file, specifically: the first node sends a write lock application request corresponding to the target file to the lock management device, so that the lock management device reclaims the target file from the second node The write lock. After the lock management device receives the write lock application request corresponding to the target file sent by the first node, it determines whether other nodes have applied for the write lock corresponding to the target file, and the lock management device finds that the second node has applied for the write lock corresponding to the target file. The second node has made the latest modification to the target file, and the lock management device sends a write lock recovery command corresponding to the target file to the second node, and the second node prepares to return the lock after receiving the write lock recovery command sent by the lock management device, and The difference in the prior art is that when the second node returns the lock, it does not offload and invalidate the data in the cache of the second node, and the data in the cache of the second node is still valid. The second node returns the write lock to the lock management device, the lock management device sends a write lock authorization response to the first node, the write lock includes the information of the second node, and the first node receives the write lock authorization response sent by the lock management device.

Step 102, the first node obtains the latest global cache table corresponding to the target file according to the information of the second node.

First, introduce the global cache table. Each embodiment of the present invention establishes a global cache table (Global Cache Table, GCT for short) for each file in the cluster, that is, the cache distribution table of the latest data in the cluster. This table records the The modification of the file in the cluster is updated to each node with the read lock and write lock. When a node modifies the file, the modification record of the node to the file is updated to the global cache table, and other When a node accesses the file, it first obtains the latest global cache table corresponding to the file, and the latest global cache table is stored on the node that has recently modified the file. In this embodiment, the target file refers to the file currently to be accessed by the client, and the first node obtains the latest global cache table corresponding to the target file according to the information of the second node. The latest modification record, for example, which data block in the target file has been modified by the second node.

The first node obtains the latest global cache table according to the information of the second node, specifically: the first node sends a global cache table synchronization request message to the second node, and after the second node receives the global cache table synchronization request message sent by the first node , return the global table table synchronization request response to the first node, the first node receives the global cache table synchronization response message sent by the second node, the global cache table synchronization response message contains the latest global cache table, and the latest global cache table stores the first The latest modification record of the target file by the second node.

Step 103, the first node modifies the target file according to the latest modification record of the second node to the target file, saves the modified data in the buffer memory of the first node, and updates the modification record of the target file by the first node to The latest global cache table.

After the first node obtains the latest global cache table, it modifies the target file according to the latest modification record of the target file recorded by the second node in the latest global cache table, specifically, the first node modifies the target file according to the second node's The latest modification record determines whether to obtain the modification data of the target file by the second node. If so, that is, the first node determines that it needs to obtain the modification data of the target file by the second node, then the first node obtains the modification data of the target file by the second node , store the modified data of the obtained target file in the cache of the first node, so that the data of the target file stored on the first node and the second node are consistent, and then, the first node reads from the cache of the first node Get the data and modify the target file. If not, that is, the data of the target file stored on the first node and the second node are consistent, the first node directly reads the data from the cache of the first node, and modifies the target file.

Generally, the target file includes at least two data blocks. When the first node and the second node modify the target file alternately, it may be to modify the same data block in the target file, or to modify different data blocks. Modification, if the same data block is modified, when the second node is modified and the first node is modified again, the data of the same data block saved on the first node at this time is not the latest, and the latest data is saved in the second In this case, the first node needs to obtain the modification data of the same data block from the second node to obtain the latest modification data of the same data block; if the data to be modified by the first node and the The data to be modified by the second node is not the same data block, so it is not necessary to obtain the modification data of the target file by the second node from the second node.

In this embodiment, the latest modification record of the target file by the second node may include: information of the data block that the second node has recently modified for the target file. The first node determines whether to obtain the modification data of the target file by the second node according to the latest modification record of the second node to the target file, specifically: the first node performs the latest modification of the target file according to the information of the second node, Determine whether the data block that the first node performs write access to the target file is the same data block as the second node that has recently modified the target file; if so, the data block that the first node performs write access to the target file is the same as The data block that the second node has recently modified on the target file is the same data block, and the first node obtains the modification data on the target file by the second node. If the data block that the first node performs write access to the target file is not the same data block as the data block that the second node has recently modified the target file, then the first node does not need to obtain the modification data of the second node to the target file, Read data directly from the cache of the first node for modification. Through this method, the data interaction between nodes is transmitted according to the needs. If different data blocks are written or read, only the latest global cache table can be exchanged between nodes, which avoids unnecessary data transmission and occupies The network overhead is low.

In this embodiment, the first node obtains the modification data of the target file by the second node according to the latest modification record of the second node to the target file, specifically: the first node obtains the latest modification data of the target file by the second node in the latest global cache table Modify the record and send a data synchronization request message to the second node. After receiving the data synchronization request message, the second node returns a data synchronization response message to the first node. The first node receives the data synchronization response message sent by the second node, and the data synchronization The response message includes modification data of the target file by the second node. Here, the data synchronization request message is used to obtain the data modified by the second node to the target file from the second node. In one mode, the first node can only obtain the data modified by the second node to the target file, such as the second node only If the data in the first data block of the target file is modified, the first node only needs to obtain the modified data corresponding to the first block, and does not need to obtain all the data of the target file, which can reduce the number of the first node and the second node transfer of data between them. Certainly, the data of the entire target file stored on the second node may also be acquired, and the present invention is not limited thereto. The first node obtains the modification data of the first data block by the second node from the second node, so that the data of the target file stored in the cache on the first node is the latest data, so that the first node and the second node The data between them can be synchronized to ensure that the data of the target file stored on each node is consistent.

Step 104, the first node writes the target file, saves the modified data in the cache of the first node, and updates the modification record of the target file by the first node to the latest global cache table.

When the first node determines that the data in the global cache table and the target file saved on the first node are all up-to-date, it reads the target file from the cache of the first node, writes the target file, and saves the modified data in In the cache of the first node, after successfully modifying the target file, the first node updates the modification record of the target file to the latest global cache table. Due to write access, the data of the target file will be modified, so each After the node modifies the target file, it must update the global cache table stored by itself.

In the method provided by this embodiment, by defining a global cache table for the target file, the global cache table contains the modification records of each node to the target file, and when the first node performs write access to the target file, it The information of the second node whose file has been newly modified is obtained from the latest global cache table, and the modification data of the second node to the target file is obtained according to the latest modification record of the second node to the target file in the latest global cache table. When two nodes In the case of interactively writing different data of the target file at the same time, the data in each cache is valid, and subsequent reads can be obtained directly from the cache. When two nodes write the same data interactively, they need to be modified from another node. Data, the transfer of data between networks is much more efficient than one-time disk writing and disk reading, so whether it is cross-modification of the same data or different data, the performance of the cluster file system has been greatly improved compared with the existing methods.

On the basis of Embodiment 1, in another embodiment of the present invention, after step 104, the following steps may also be included: when the first node performs read access to the target file, apply to the lock management device for a read lock corresponding to the target file , the read lock includes information about the first node that has recently modified the target file; the first node reads the target file from the cache of the first node. After the first node modifies the target file, if the first node applies for read access to the target file again, the first node first applies to the lock management device for a read lock corresponding to the target file, and the lock management device receives the first node's send After the read lock application request, the cache on the first node is not written to the disk, the data in the cache of the first node is still valid, and the lock management device authorizes the read lock to the first node. The information of the first node where the file has been newly modified. The first node finds that it is the node that has made the latest modification to the target file according to the read lock. Therefore, the data of the target file stored in the cache of the first node is valid. A node reads the object file from the cache.

On the basis of Embodiment 1, in yet another embodiment of the present invention, after step 104, the following steps may also be included: the first node receives the global cache table synchronization request message sent by the second node, and the global cache table synchronization request message It is sent to the first node when the second node needs to perform read or write access to the target file. Specifically, after the first node modifies the target file, if the second node needs to perform read or write access to the target file, the second node first applies to the lock management device for a read lock or write lock corresponding to the target file, and the lock After the management device receives the read lock or write lock application request sent by the second node, it does not write to the cache on the first node, and the data in the cache of the first node is still valid, and the lock management device will read or write the lock. The write lock is authorized to the second node, and the read lock or write lock contains the information of the first node that has made the latest modification to the target file, and the second node sends a global cache table synchronization request message to the first node according to the information of the first node After receiving the global buffer table synchronization request message, the first node returns a global cache table synchronization response message to the second node, and the global cache table synchronization response message includes the updated latest global cache table according to the updated latest global cache table.

After obtaining the updated latest global cache table, the second node takes the updated global cache table as the latest global cache table, and determines the first node according to the modification record of the target file in the latest global cache table. Whether the data modified by a node to the target file is the same as the data of the target file to be accessed by itself, if they are the same, it means that the data saved on the first node is the latest, and the second node sends a data synchronization request to the first node, the first The node receives the data synchronization request sent by the second node, and the data synchronization request message is sent by the second node according to the modification record of the first node to the target file in the updated latest global cache table.

The first node sends a data synchronization response to the second node, and the data synchronization response includes the modification data of the first node to the target file, so that the second node saves the modification data of the first node to the target file in the cache of the second node , when the second node performs read access or write access to the data, the target file may be read from the cache of the second node.

FIG. 4 is a flowchart of Embodiment 2 of the cache synchronization method of the distributed cluster file system of the present invention. As shown in FIG. 4, the method provided in this embodiment includes the following steps:

Step 201. The lock management device receives a write lock application request corresponding to the target file sent by the first node. The write lock application request is sent by the first node to the lock management device when the client performs write access to the target file.

When the client performs write access to the target file, the first node first applies to the lock management device for a write lock corresponding to the target file.

Step 202, the lock management device determines that the second node has applied for the write lock corresponding to the target file, and has made the latest modification to the target file, then the lock management device sends the write lock recovery command corresponding to the target file to the second node, and reclaims the write lock to the second node. For the write lock corresponding to the target file, when the lock management device reclaims the write lock corresponding to the target file from the second node, it does not write and invalidate the data of the target file in the cache of the second node.

After receiving the write lock application request corresponding to the target file sent by the first node, the lock management device determines whether other nodes have applied for the write lock corresponding to the target file. If the lock management device determines that the second node has applied for the write lock corresponding to the target file For a write lock, if the target file has been newly modified, the lock management device reclaims the write lock corresponding to the target file from the second node. Different from the prior art, when the lock management device reclaims the write lock corresponding to the target file from the second node, it does not write and invalidate the data of the target file in the cache of the second node. Therefore, the cache of the second node The stored data of the target file is still valid, and if the second node accesses the target file again later, it does not need to read from the disk, but directly reads the required data in the target file from the cache.

Step 203, the lock management device sends a write lock authorization response to the first node, and the write lock includes information of the second node.

After the lock management device reclaims the write lock corresponding to the target file from the second node, it sends a write lock authorization response to the first node, authorizes the write lock corresponding to the target file to the first node, and includes the write lock on the target file in the write lock. Information about the last modified second node. After the first node obtains the write lock, it finds that the second node has applied the write lock to the target file, then the first node determines that the data of the target file stored in its own cache is not up-to-date. The latest global cache table is obtained at the location, and the latest modification record of the target file by the second node is stored in the latest global cache table. The first node determines whether to acquire modification data of the target file by the second node according to the latest modification record of the target file by the second node.

In the method provided in this embodiment, when the first node applies to the lock management device for the write lock corresponding to the target file, the lock management device determines that the second node has also applied for the write lock corresponding to the target file, and the lock management device recovers the lock from the second node. When the write lock corresponding to the target file is applied, the data in the cache of the second node is not written and invalidated, and the data in the cache of the second node is still valid, so that when the second node modifies the target file again, it can be accessed from the second node. The data is read in the cache of the two nodes, even if the first node and the second node modify the target file alternately, the data in the caches of the two nodes are valid, avoiding frequent disk reading and writing.

After the lock management device in this embodiment sends a write lock authorization response to the first node, if the first node needs to perform read access to the target file, the method provided in this embodiment may further include the following steps: the lock management device receives the The lock management device determines that the first node has applied for the write lock corresponding to the target file, and the target file has been updated, and the lock management device sends the write lock recovery corresponding to the target file to the first node. command to reclaim the write lock corresponding to the target file to the first node, and the lock management device does not write and invalidate the data of the target file in the cache of the first node when reclaiming the write lock corresponding to the target file from the first node, so , the data in the cache of the first node is still valid. After the lock management device reclaims the write lock corresponding to the target file from the first node, it sends a read lock authorization response to the first node, and the read lock includes the information of the first node, so that the first node reads from the cache of the first node Target file. Here, since the lock management device reclaims the write lock corresponding to the target file from the first node, it does not write and invalidate the data in the cache of the first node, the data in the cache of the first node is still valid, and the first node can Read data directly from the cache. If the subsequent second node also needs to read the data of the target file, it can obtain the modification data of the first node to the target file according to the modification record of the first node to the target file. The first node will The modified data of the target file is sent to the second node, so that the data of the two nodes are consistent, thus avoiding the second node from reading data from the disk, because the transmission of data between networks is more efficient than writing and reading disk once Many, improving resource utilization.

After the lock management device in this embodiment sends a write lock authorization response to the first node, if the second node needs to perform read or write access to the target file, the method provided in this embodiment may further include the following steps: the lock management device receives the first The read lock or write lock application request corresponding to the target file sent by the second node, the lock management device determines whether other nodes have applied for the target file after receiving the read lock or write lock application request corresponding to the target file sent by the second node For the corresponding write lock or read lock, the lock management device determines that the first node has applied for the write lock corresponding to the target file, and the target file has been newly modified, then the lock management device sends a write lock recovery command corresponding to the target file to the first node, Reclaim the write lock corresponding to the target file to the first node. When the lock management device reclaims the write lock corresponding to the target file from the first node, it does not write and invalidate the data of the target file in the cache of the first node. The first node The data in the cache is still valid. After the lock management device reclaims the write lock corresponding to the target file from the first node, it sends a read lock or write lock authorization response to the second node. The read lock or write lock includes the first node's Information, so that the second node obtains the latest global cache table according to the information of the first node. After the second node obtains the authorization of the read lock or write lock and learns that the first node has modified the target file, the read lock or write lock includes the information of the first node, and obtains the latest global cache table from the first node. The modification record of the first node to the target file is recorded in the global cache table. And according to the modification record of the first node to the target file, it is determined whether it is necessary to synchronize data from the first node. Specifically, the first node determines the data block that the first node performs write access to the target file and the data block to be accessed by the second node. Whether it is the same data block, for example, if the second node wants to access the 0th data block in the target file, the first node has modified the 2nd data block of the target file, but the first node has not modified the 0th data block The data block has been modified, and the data of the 0th data block of the target file saved on the second node is still the latest, then the second node does not need to obtain the modified data of the target file from the first node, and reads it from the cache of the second node If there is no data in the cache, read it from the disk of the second node. If the second data is to be accessed by the second node, the second node needs to obtain the modification data of the target file by the first node, so that the data on the two nodes are synchronized. The second node transmits the data from the first node through the network, thereby avoiding the second node from reading data from the disk. Since the data transmission between the networks is much more efficient than writing and reading disks once, resource utilization is improved. utilization rate.

FIG. 5 is a flow chart of Embodiment 3 of the cache synchronization method of the distributed cluster file system of the present invention. As shown in FIG. 5, the method provided in this embodiment includes the following steps:

Step 301, the second node receives the global cache table synchronization request message sent by the first node.

After receiving the write lock or read lock authorization response sent by the lock management device, the first node sends a global cache table synchronization request message to the second node for obtaining the latest global cache table.

Step 302, the second node returns a global cache table synchronization response message to the first node according to the global cache table synchronization request message, the global cache table synchronization response message includes the latest global cache table, and the latest global cache table stores the second node pair A record of the last modification of the object file.

After the second node sends the latest global cache table to the first node, the latest modification record of the second node to the target file is stored in the latest global cache table, and the first node determines whether to It is necessary to obtain the modification data of the target file by the first node. If necessary, the first node sends a data synchronization request message to the second node, and the second node receives the data synchronization request message sent by the first node; the second node receives the data synchronization request message according to the data synchronization request message, returning a data synchronization response message to the first node, where the data synchronization response message includes the modification data of the target file by the second node. Thus, the data on the first node and the second node are synchronized.

FIG. 6 is a flow chart of Embodiment 4 of the cache synchronization method of the distributed cluster file system of the present invention. This embodiment will describe the scheme shown in Embodiment 1 in detail through a specific example. As shown in FIG. 6 , this embodiment The method that embodiment provides comprises the following steps:

Step 401. When the second node wants to write the fourth data block of file 1, the second node sends a write lock application request corresponding to file 1 to the lock management device.

In this embodiment, the file 1 is the target file, and the file 1 includes six data blocks, the 0th data block, the 1st data block, the 2nd data block, the 3rd data block, the 4th data block, and the 5th data block. When both the first node and the second node do not write to file 1, the global cache table corresponding to file 1 is empty. When the global cache table is empty, read access is performed, and the data is read from the cache of the node first. If If there is data to be read in the cache of the node, it returns from the cache. If there is no data to be read in the cache of the node, it is read from the disk. Afterwards, the second node performs write access to the file 1 to write the fourth data block of the file 1, and the second node first applies for a write lock corresponding to the file 1 from the lock management device.

In step 402, the lock management device returns a write lock application response corresponding to file 1 to the second node.

After receiving the write lock application request sent by the second node, the lock management device determines that no other node has applied for the write lock corresponding to file 1, and authorizes the write lock corresponding to file 1 to the second node.

Step 403, the second node modifies the fourth data block of file 1, and updates the global cache table corresponding to file 1.

After the second node holds the write lock of file 1, it modifies the fourth data block of file 1, saves the modified data in the cache, and updates the global cache table corresponding to file 1 at the same time. Add a new record item to the table, each file corresponds to a global cache table, there are multiple record items in the global cache table, each record item corresponds to a data block, if there are 5 data blocks in the file, each data block A record item is established, and when a node modifies a data block, the record item corresponding to the data block in the global cache table is updated. In this step, since the global cache table was empty before, a new record item needs to be added, which is the record item corresponding to the fourth data block. Here, the latest modification record of the second node to the fourth data block of file 1 is recorded Update to the record item corresponding to the 4th data block.

Step 404, when the first node wants to write the fourth data block of file 1, the first node applies for a write lock corresponding to file 1 from the lock management device.

Step 405, the lock management device sends a write lock recovery request corresponding to file 1 to the second node.

Step 406, the second node returns a write lock recovery response corresponding to file 1 to the lock management device.

After the second node receives the write lock recovery request corresponding to file 1 sent by the lock management device, it prepares to return the lock. When the lock is returned, it does not write and invalidate the data in the second node cache, and the data in the second node cache Data is still valid. Step 407, the lock management device returns a write lock application response corresponding to file 1 to the first node.

The lock management device authorizes the write lock corresponding to the file 1 to the first node, and the write lock includes the information of the second node that has made the latest modification to the file 1 .

Step 408, the first node sends a global cache table synchronization request message to the second node.

The first node sends a global cache table synchronization request message to the second node to obtain the latest global cache table from the second node.

Step 409, the global cache table synchronization response message sent by the second node to the first node.

The global cache table synchronization response message includes the latest global cache table, and the latest global cache table stores the latest modification record of the second node to the file 1, that is, the modification record of the second node to the fourth data block of the file 1.

Step 410, the first node sends a data synchronization request message of the fourth data block to the second node according to the latest modification record of the second node to the target file in the latest global cache table.

After the first node receives the global cache table synchronization response message sent by the second node, it determines that the second node has modified the fourth data block of file 1 according to the latest global cache table of file 1, and then the first node sends data to the second node The synchronization request message is used to obtain the modification data of the fourth data block of the file 1 by the second node.

Step 411, the second node sends a data synchronization response message of the fourth data block to the first node.

Wherein, the data synchronization response message of the fourth data block includes the modification data of the second node to the fourth data block of the target file.

Step 412, the first node modifies the fourth data block, saves the modified data in the cache of the first node, and updates the global cache table corresponding to file 1.

After the first node receives the modification data of the fourth data block sent by the second node, it fills the modification data of the fourth data block into the pre-allocated cache, completes the data synchronization of the two nodes, and then performs the real modification. Save the modified data of the fourth data block in the cache. Since the modification record item of the fourth data block has been added to the global cache table when the second node modifies the fourth data block, here the first node updates the record of the global cache table after the modification of the fourth data block is completed item, that is, the update is: the fourth data block of file 1 is modified by the first node.

Step 413 , when the first node wants to modify the third data block of file 1, query the global cache table, determine that no other node has modified the third data block, and modify the third data block.

If the first node modifies the third data block of file 1 again, queries the global cache table, and finds that no other node has modified the third data block, no need to synchronize, just modify it directly, and the third data block is not in the global cache table For related record items, the record item of the third data block needs to be added, and the third data block of file 1 has been modified by the first node.

Step 414: When the second node wants to read the fourth data block of file 1, it sends a read lock application request corresponding to file 1 to the lock management device.

If the subsequent second node wants to perform read access to the fourth data block of file 1, first apply for the read lock corresponding to file 1,

Step 415, the lock management device returns a read lock application response corresponding to file 1 to the second node.

Step 416, the second node sends a global cache table synchronization request to the first node.

After the second node obtains the read lock, the read lock contains the information of the first node that has recently modified the fourth data block of file 1. The first node has applied for a write lock before, so the global cache table on the first node up to date. The second node initiates a global cache table synchronization request to the first node.

Step 417, the first node returns a global cache table synchronization response to the second node.

The global cache table synchronization response includes the latest global cache table. After obtaining the latest global cache table, the second node determines that the fourth data block has been modified by the first node, and then initiates a data synchronization request to the first node.

Step 418, the second node sends a data synchronization request of the fourth data block to the first node.

Step 419, the first node returns a data synchronization response of the fourth data block to the second node.

The data synchronization response of the 4th data block includes the modified data of the 4th data block by the first node, and the second node obtains the modified data of the 4th data block by the first node, so that the data on the two nodes are synchronized, and The modified data is stored in the cache of the second node.

Step 420, the second node reads the data of the fourth data block from the cache of the second node.

In this embodiment, the first node and the second node may also write or fetch file 1, which will not be listed one by one here. For example, the second node also needs to read the second data block. When reading, it finds that the first 2. The data block is modified by the second node, so the cache of the second node must be the latest data, just return it from the cache. If the second node reads the 5th data block and finds that the 5th data block has not been modified by any node, follow the normal reading process: if it exists in the cache, return from the cache, if not, read it from the disk.

It should be noted that, in each embodiment of the present invention, read access can be concurrent. The second node performs write access, and other nodes all perform read access to it. If other nodes access the data modified by the second node, all The corresponding data is requested from the second node.

In the method provided by this embodiment, by establishing a global cache table for files in the cluster, this table is migrated to the node currently accessing the file along with the lock authorization, and when accessing, the global cache table is queried to determine whether the accessed data is It is modified by other nodes and determines whether data needs to be synchronized. It uses less overhead to achieve "on-demand" data synchronization. After adding GCT, it is not necessary to cache the latest data of the file on each node, but only cache the data that needs to be used. It will greatly reduce the consumption of the cache and avoid frequent disk reads and writes, so as to improve the performance of the cluster file system.

The method provided in this embodiment can avoid the no-cache operation caused by frequent writing and reading of the same file when cluster nodes concurrently and alternately access the same file. At the same time, the method provided in this embodiment implements a "on-demand" interactive data solution. The data interaction between nodes is transmitted according to the needs. In most cases, only the global cache table can be exchanged between nodes, which avoids unnecessary data transmission, occupies less network overhead, and saves data between networks. Transferring is much more efficient than writing to disk → reading disk once. In the case of interactively writing different data at the same time, each cache is valid, and subsequent reads can be obtained directly from the cache. Therefore, whether it is cross-modification of the same data or different data, the cluster file Compared with the existing methods, the system performance has been greatly improved.

For the convenience of description, only two nodes are taken as an example in each embodiment of the present invention, but the present invention can be applied to a cluster of n nodes (n>=2), and the present invention does not limit the number of nodes , when multiple nodes alternately perform read and write access to the target file at the same time, the solutions provided by the embodiments of the present invention are still applicable.

FIG. 7 is a schematic structural diagram of the first node of the present invention, including: a lock application module 51 , a cache table acquisition module 52 , and a file modification module 53 .

The lock application module 51 is used to apply to the lock management device for a write lock corresponding to the target file when the client performs write access to the target file on the first node, and the write lock includes the second node that has carried out the latest modification to the target file Information;

The cache table obtaining module 52 is used to obtain the latest global cache table corresponding to the target file according to the information of the second node, and the latest modification record of the second node to the target file is stored in the latest global cache table;

The file modification module 53 is used to modify the target file according to the latest modification record of the second node to the target file, save the modified data in the cache of the first node, and update the modification record of the target file by the first node to the latest global cache table.

The lock application module 51 is specifically configured to: send a write lock application request corresponding to the target file to the lock management device, so that the lock management device reclaims the write lock to the second node; receive a write lock authorization response sent by the lock management device.

File modification module 53 comprises judgment unit, data acquisition unit and file modification unit; Judgment unit is used for determining whether to obtain the modification data of second node to target file according to the latest modification record of second node to target file; If so, then acquisition unit Obtain the modification data of the target file by the second node, store the modified data of the obtained target file in the cache of the first node, so that the data of the target file stored on the first node and the second node are consistent, and the file modification unit Read data from the cache of the first node to modify the target file; if not, the file modification unit reads data from the cache of the first node to modify the target file.

In this embodiment, the target file includes at least two data blocks, and the latest modification record of the second node to the target file includes: the information of the data block that the second node has recently modified the target file; the judging unit is specifically used to: The second node has carried out the information of the data block that the target file has been modified recently, and it is determined whether the data block that the first node has performed write access to the target file and the data block that the second node has carried out the latest modification to the target file are the same data block; If yes, the data obtaining unit obtains the modification data of the target file by the second node.

The cache table obtaining module 52 specifically obtains the latest global cache table in the following manner: sending a global cache table synchronization request message to the second node; receiving a global cache table synchronization response message sent by the second node, the global cache table synchronization response message contains the latest global cache table In the cache table, the latest modification record of the target file by the second node is stored in the latest global cache table.

The data acquisition unit is specifically used to obtain the modified data in the following manner: according to the latest modification record of the second node to the target file in the latest global cache table, send a data synchronization request message to the second node; receive a data synchronization response message sent by the second node , the data synchronization response message includes the modification data of the target file by the second node.

The first node provided in this embodiment can be used to implement the technical solutions of the method embodiments 1 and 4, and the specific implementation manner and technical effect are similar and will not be repeated here.

Fig. 8 is a schematic structural diagram of Embodiment 2 of the first node of the present invention. As shown in Fig. 8, the first node of this embodiment may further include: file reading on the basis of the structure of the first node shown in Fig. 7 Fetching module 54, receiving module 55, sending module 56.

Wherein, the lock application module 51 is also used for: when the first node performs read access to the target file, apply to the lock management device for the read lock corresponding to the target file, and the read lock includes the first node that has carried out the latest modification to the target file. information;

The file reading module 54 is configured to read the target file from the cache of the first node.

After the file modification module 53 modifies the target file, if the receiving module 55 receives the global cache table synchronization request message sent by the second node, the global cache table synchronization request message is that the second node needs to read or write the target file. If it is sent to the first node during access, the sending module 56 is used to return the global cache table synchronization response message to the second node according to the global cache table synchronization request message, and the global cache table synchronization response message includes the updated latest global cache surface.

The receiving module 55 is also used to: receive the data synchronization request sent by the second node, the data synchronization request message is sent by the second node according to the modification record of the first node in the updated latest global cache table to the target file; the sending module 56 is also Used to: send a data synchronization response to the second node, the data synchronization response includes the modification data of the target file by the first node, so that the second node saves the modification data of the target file by the first node in the cache of the second node , when the second node performs read access or write access to the data, the target file may be read from the cache of the second node.

The first node provided in this embodiment can be used to implement the technical solutions of the method embodiments 1 and 4, and the specific implementation manner and technical effect are similar and will not be repeated here.

FIG. 9 is a schematic structural diagram of Embodiment 1 of the lock management device of the present invention. The lock management device of this embodiment includes: a receiving module 61 , a determining module 62 , and a sending module 63 .

Wherein, the receiving module 61 is configured to receive the write lock application request corresponding to the target file sent by the first node, and the write lock application request is sent by the first node to the lock management device when the client performs write access to the target file;

A determining module 62, configured to determine that the second node has applied for a write lock corresponding to the target file, and the second node has carried out the latest modification to the target file;

The sending module 63 is configured to send a write lock recovery command corresponding to the target file to the second node, and recover the write lock corresponding to the target file from the second node, and when the lock management device recovers the write lock corresponding to the target file from the second node, and Do not write to disk and invalidate the data of the target file in the cache of the second node;

The sending module 63 is further configured to: send a write lock authorization response to the first node after the lock management device reclaims the write lock corresponding to the target file from the second node, and the write lock includes information of the second node.

The receiving module 61 is also used for: receiving the read lock application request corresponding to the target file sent by the first node; the determining module 62 is also used for: determining that the first node has applied for the write lock corresponding to the target file, and the first node has applied for the target file. The latest modification; the sending module 63 is also used to: send the write lock reclaim command corresponding to the target file to the first node, reclaim the write lock corresponding to the target file to the first node, and the lock management device reclaims the write lock corresponding to the target file to the first node When locking, the data of the target file in the cache of the first node is not written to disk and invalid; the sending module 63 is also used to: send a read lock authorization response to the first node, and the read lock includes the information of the first node, so that The first node reads the target file from the cache of the first node.

In this embodiment, the receiving module 61 is also used to: receive the read lock or write lock application request corresponding to the target file sent by the second node; the determination module 62 is also used to: determine that the first node has applied for the write lock corresponding to the target file, The first node has carried out the latest modification to the target file; the sending module 63 is also used to: send the write lock reclaim command corresponding to the target file to the first node, reclaim the write lock corresponding to the target file to the first node, and the lock management device is sending to the first node When a node reclaims the write lock corresponding to the target file, it does not write and invalidate the data of the target file in the cache of the first node; the sending module 63 is also used to: send a read lock or write lock authorization response to the second node, read The lock or write lock includes the information of the first node, so that the second node obtains the latest global cache table according to the information of the first node.

The lock management device provided in this embodiment can be used to implement the technical solution provided in the second method embodiment, and the specific implementation method and technical effect are similar, and will not be repeated here.

FIG. 10 is a schematic structural diagram of Embodiment 1 of the second node of the present invention. The second node provided in this embodiment includes: a receiving module 71 and a sending module 72 .

Wherein, the receiving module 71 is configured to receive the global cache table synchronization request message sent by the first node;

The sending module 72 is configured to return a global cache table synchronization response message to the first node according to the global cache table synchronization request message, the global cache table synchronization response message includes the latest global cache table, and the latest global cache table stores the second node pair A record of the last modification of the object file.

The receiving module 71 is also used for: receiving the data synchronization request message sent by the first node, the data synchronization request message is sent by the first node according to the latest modification record of the second node in the latest global cache table to the target file; the sending module 72 also uses In: returning a data synchronization response message to the first node according to the data synchronization request message, where the data synchronization response message includes modification data of the target file by the second node.

The second node provided in this embodiment can be used to execute the technical solution provided in the third embodiment of the method, and the specific implementation method and technical effect are similar, and will not be repeated here.

Fig. 11 is a schematic structural diagram of the third embodiment of the first node of the present invention. As shown in Fig. 11, the first node 800 of this embodiment includes: a processor 81, a memory 82, a transmitter 83 and a receiver 84, a memory 82, a transmitter The processor 83, the receiver 84 and the processor 81 are connected through the bus, and the memory 82 stores execution instructions. When the first node 800 is running, the processor 81 communicates with the memory 82, and the processor 81 executes the instructions so that the first node 800 executes the present In the cache synchronization method of the distributed cluster file system provided by the invention, the processor 81 is specifically used to perform the following operations:

When the client performs write access to the target file on the first node, it applies to the lock management device for a write lock corresponding to the target file, and the write lock includes the information of the second node that has carried out the latest modification to the target file;

According to the information of the second node, the latest global cache table corresponding to the target file is obtained, and the latest modification record of the second node to the target file is stored in the latest global cache table;

The target file is modified according to the latest modification record of the second node to the target file, the modified data is stored in the cache of the first node, and the modification record of the first node to the target file is updated to the latest global cache table.

The processor 81 applies to the lock management device for a write lock corresponding to the target file, specifically: the processor 81 sends a write lock application request corresponding to the target file to the lock management device through the transmitter 83, so that the lock management device reclaims the write lock to the second node. lock, and receive the write lock authorization response sent by the lock management device through the receiver 84.

The processor 81 modifies the target file according to the latest modification record of the second node to the target file, specifically: determining whether to obtain the modification data of the second node to the target file according to the latest modification record of the second node to the target file; if so, then Obtain the modification data of the target file from the second node, and store the obtained modification data of the target file in the cache of the first node, so that the data of the target file stored on the first node and the second node are consistent, from the first The data is read from the cache of the node, and the target file is modified; if not, the data is read from the cache of the first node, and the target file is modified.

In a possible implementation of the present invention, the target file includes at least two data blocks, and the latest modification record of the target file by the second node includes: information of the data block that the second node has recently modified for the target file;

The processor 81 determines whether to obtain the modification data of the target file by the second node according to the latest modification record of the target file by the second node, specifically: according to the information of the data block that has been newly modified by the second node to the target file, determine the first Whether the data block that the node performs write access to the target file and the data block that the second node has recently modified on the target file are the same data block; if so, obtain the modification data of the target file by the second node.

The first node obtains the latest global cache table according to the information of the second node, specifically:

The transmitter 83 sends a global cache table synchronization request message to the second node, and then the receiver 84 receives the global cache table synchronization response message sent by the second node, the global cache table synchronization response message includes the latest global cache table, the latest global cache table The latest modification record of the second node to the target file is saved in .

The processor 81 obtains the modification data of the second node to the target file according to the latest modification record of the second node to the target file, specifically:

The processor 81 sends a data synchronization request message to the second node through the transmitter 83 according to the latest modification record of the second node in the latest global cache table to the target file; then, the receiver 84 receives the data synchronization response message sent by the second node, The data synchronization response message includes the modification data of the target file by the second node.

In this embodiment, when the first node writes the target file, saves the modified data in the cache of the first node, and updates the modification record of the target file by the first node to the latest global cache table, the processing The device 81 is also used for: when the first node performs read access to the target file, apply to the lock management device for the read lock corresponding to the target file, and the read lock includes the information of the first node that has carried out the latest modification to the target file. After the write lock is granted, the target file is read from the cache of the first node.

In this embodiment, the first node performs a write operation on the target file, saves the modified data in the cache of the first node, and updates the modification record of the target file by the first node to the latest global cache table, the processor 81 Also used for:

The global cache table synchronization request message sent by the second node is received by the receiver 84. The global cache table synchronization request message is sent to the first node when the second node needs to perform read access or write access to the target file. Then, the processor 81 According to the global cache table synchronization request message, the transmitter 83 returns a global cache table synchronization response message to the second node, and the global cache table synchronization response message includes the latest updated global cache table.

In this embodiment, after the first node returns the global cache table synchronization response message to the second node or the third node according to the global cache table synchronization request message, it further includes:

The receiver 84 receives the data synchronization request sent by the second node, and the data synchronization request message is sent by the second node according to the modification record of the first node in the updated latest global cache table to the target file; the transmitter 83 sends to the second node Data synchronization response, the data synchronization response includes the modification data of the first node to the target file, so that the second node saves the modification data of the first node to the target file in the cache of the second node, when the second node performs the modification on the data The target file may be read from the cache of the second node during read access or write access.

The first node provided in this embodiment can be used to implement the technical solutions provided in the first and fourth method embodiments, and the specific implementation and technical effects are similar, and will not be repeated here.

Fig. 12 is a schematic structural diagram of Embodiment 2 of the lock management device of the present invention. As shown in Fig. The processor 93 and the receiver 94 are connected to the processor 91 through a bus, and the memory 92 stores execution instructions. When the lock management device 900 is running, the processor 91 communicates with the memory 92, and the processor 91 executes the instructions.

Wherein, the receiver 94 is configured to receive the write lock application request corresponding to the target file sent by the first node, and the write lock application request is sent by the first node to the lock management device when the client performs write access to the target file;

The processor 91 is configured to determine that the second node has applied for a write lock corresponding to the target file, and the target file has been newly modified, then the lock management device sends a write lock reclaim command corresponding to the target file to the second node, and reclaims the write lock to the second node. For the write lock corresponding to the target file, when the lock management device reclaims the write lock corresponding to the target file from the second node, it does not write and invalidate the data of the target file in the cache of the second node;

The transmitter 93 is configured to send a write lock authorization response to the first node, where the write lock includes information of the second node.

After the transmitter 93 sends the write lock authorization response to the first node, the receiver 94 is also used to receive the read lock application request corresponding to the target file sent by the first node; the processor 91 is also used to determine that the first node has applied for the target file corresponding to If the target file has been newly modified, the lock management device sends a write lock recovery command corresponding to the target file to the first node, and reclaims the write lock corresponding to the target file to the first node. When the write lock corresponding to the target file is reclaimed, the data of the target file in the cache of the first node is not written to disk and invalidated; the transmitter 93 lock management device sends a read lock authorization response to the first node, and the read lock includes the first node information, so that the first node reads the target file from the cache of the first node.

The lock management device provided in this embodiment can be used to implement the technical solution provided in the second method embodiment, and the specific implementation method and technical effect are similar, and will not be repeated here.

Fig. 13 is a schematic structural diagram of Embodiment 2 of the second node of the present invention. As shown in Fig. 13, the second node 100 of this embodiment includes: a processor 11, a memory 12, a transmitter 13 and a receiver 14, a memory 12, a transmitter The processor 13 and the receiver 14 are connected to the processor 11 through a bus, and the memory 12 stores execution instructions. When the second node 100 is running, the processor 11 communicates with the memory 12, and the processor 11 executes the instructions.

In this implementation, the receiver 14 is configured to receive the global cache table synchronization request message sent by the first node;

The processor 11 is configured to return a global cache table synchronization response message to the first node through the transmitter 13 according to the global cache table synchronization request message, the global cache table synchronization response message includes the latest global cache table, and the latest global cache table stores the The latest modification record of the target file by the second node.

After the transmitter 13 returns the global cache table synchronization response message to the first node, the receiver 14 is also used to: receive the data synchronization request message sent by the first node, the data synchronization request message is the first node according to the first node in the latest global cache table. Two nodes send the latest modification record of the target file; the processor 11 returns a data synchronization response message to the first node through the transmitter 13 according to the second node according to the data synchronization request message, and the data synchronization response message includes the second node's request for the target file. Modification data of the file.

The second node provided in this embodiment can be used to execute the technical solution provided in the third embodiment of the method, and the specific implementation method and technical effect are similar, and will not be repeated here.

Those of ordinary skill in the art can understand that all or part of the steps for implementing the above method embodiments can be completed by program instructions and related hardware. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps including the above-mentioned method embodiments; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk and other various media that can store program codes.

Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention, rather than limiting them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: It is still possible to modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the technical solutions of the various embodiments of the present invention. scope.

Claims (28)

1. a buffer memory synchronous method for distributed cluster file system, is characterized in that, comprising:

When client is carried out write access to the file destination on first node, described first node to file destination described in spot lock management device application corresponding write lock, described in write lock and comprise the information of described file destination being carried out to the Section Point of up-to-date modification;

The up-to-date global buffer table that described first node is corresponding according to file destination described in the acquisition of information of described Section Point, preserves the up-to-date amendment record of described Section Point to described file destination in described up-to-date global buffer table;

Described first node is modified to described file destination to the up-to-date amendment record of described file destination according to described Section Point, amended data are kept in the buffer memory of described first node, and described first node is updated to described up-to-date global buffer table to the amendment record of described file destination.

2. method according to claim 1, is characterized in that, described first node is modified to described file destination to the up-to-date amendment record of described file destination according to described Section Point, comprising:

Described first node determines whether to obtain the Update Table of described Section Point to described file destination according to described Section Point to the up-to-date amendment record of described file destination;

If, described first node obtains the Update Table of described Section Point to described file destination, the Update Table of the described file destination getting is stored in the buffer memory of described first node, the data of the described file destination stored on described first node and described Section Point are consistent, from the buffer memory of described first node, read described data, described file destination is modified;

If not, described first node reading out data from the buffer memory of described first node, modifies to described file destination.

3. method according to claim 2, it is characterized in that, described file destination comprises at least two data blocks, and described Section Point comprises the up-to-date amendment record of described file destination: described Section Point carried out the information of the data block of up-to-date modification to described file destination;

Described first node determines whether to obtain the Update Table of described Section Point to described file destination according to described Section Point to the up-to-date amendment record of described file destination, comprising:

Described first node carried out the information of the data block of up-to-date modification to described file destination according to described Section Point, determine that described first node carries out the data block of write access to described file destination and whether data block that described Section Point carried out up-to-date modification to described file destination is same data block;

If so, described first node obtains the Update Table of described Section Point to described file destination.

4. method according to claim 1, is characterized in that, described first node, according to the up-to-date global buffer table of the acquisition of information of described Section Point, comprising:

Described first node sends global buffer table synchronization request message to described Section Point;

Described first node receives the global buffer table synchronous response message that described Section Point sends, in described global buffer table synchronous response message, comprise described up-to-date global buffer table, in described up-to-date global buffer table, preserve the up-to-date amendment record of described Section Point to described file destination.

5. method according to claim 3, is characterized in that, described first node obtains the Update Table of described Section Point to described file destination according to described Section Point to the up-to-date amendment record of described file destination, comprising:

Described first node, according to the up-to-date amendment record of Section Point to described file destination described in described up-to-date global buffer table, sends data synchronization request message to described Section Point;

Described first node receives the data sync response message that described Section Point sends, and comprises the Update Table of described Section Point to described file destination in described data sync response message.

6. method according to claim 1, it is characterized in that, described first node carries out write operation to described file destination, amended data are kept in the buffer memory of described first node, and after described first node is updated to described up-to-date global buffer table to the amendment record of described file destination, also comprise:

When described first node carries out read access to described file destination, to read lock corresponding to file destination described in described spot lock management device application, in described read lock, comprise the information of described file destination being carried out to the described first node of up-to-date modification;

Described first node reads described file destination from the buffer memory of described first node.

7. method according to claim 1, it is characterized in that, described in described first node, file destination carries out write operation, amended data are kept in the buffer memory of described first node, and after described first node is updated to described up-to-date global buffer table to the amendment record of described file destination, also comprise:

Described first node receives the global buffer table synchronization request message that described Section Point sends, and described global buffer table synchronization request message is that described Section Point sends to described first node need to carry out read access or write access to described file destination time;

Described first node, according to described global buffer table synchronization request message, returns to global buffer table synchronous response message to described Section Point, comprises the up-to-date global buffer table after described renewal in described global buffer table synchronous response message.

8. method according to claim 7, is characterized in that, described first node is according to described global buffer table synchronization request message, after described Section Point or described the 3rd node return to global buffer table synchronous response message, also comprises:

Described first node receives the data synchronization request that described Section Point sends, and described data synchronization request message is that described Section Point sends the amendment record of described file destination according to first node described in the up-to-date global buffer table after described renewal;

Described first node sends data sync response to described Section Point, described data sync response comprises the Update Table of described first node to described file destination, so that described Section Point is kept at described first node in the buffer memory of described Section Point to the Update Table of described file destination, when carrying out read access or write access to described data, described Section Point can from the buffer memory of described Section Point, read described file destination.

9. method according to claim 1, is characterized in that, described first node to file destination described in spot lock management device application corresponding write lock, comprising:

Described first node sends to described spot lock management device the lock application request of writing that described file destination is corresponding, so that described spot lock management device is write lock described in described Section Point recovery;

Described first node receives that described spot lock management device sends writes lock authorization response.

10. a buffer memory synchronous method for distributed type assemblies target file system, is characterized in that, comprising:

Spot lock management device receives the lock application request of writing corresponding to file destination that described first node sends, described in to write lock application request be client when described file destination is carried out to write access, described first node sends to described spot lock management device;

Described spot lock management device determine Section Point application cross described file destination corresponding write lock, described file destination was carried out to up-to-date modification, described spot lock management device sends to described Section Point the lock of writing that described file destination is corresponding and reclaims order, to described Section Point reclaim described file destination corresponding write lock, described spot lock management device to described Section Point, reclaim described file destination corresponding write when lock, the data of the described file destination in the buffer memory of described Section Point are not write to dish and invalid;

Described spot lock management device sends and writes lock authorization response to described first node, described in write the information that lock comprises described Section Point.

11. methods according to claim 10, is characterized in that, described spot lock management device sends and writes lock authorization response to described first node, described in write the information that lock comprises described Section Point after, also comprise:

Described spot lock management device receives read lock application request corresponding to described file destination that described first node sends;

Described spot lock management device determine described first node application cross described file destination corresponding write lock, described file destination was carried out to up-to-date modification, described spot lock management device sends to described first node the lock of writing that described file destination is corresponding and reclaims order, to described first node reclaim described file destination corresponding write lock, described spot lock management device to described first node, reclaim described file destination corresponding write when lock, the data of the described file destination in the buffer memory of described first node are not write to dish and invalid;

Described spot lock management device sends read lock authorization response to described first node, and described read lock comprises the information of described first node, so that described first node reads described file destination from the buffer memory of described first node.

12. methods according to claim 10, is characterized in that, described spot lock management device sends and writes lock authorization response to described first node, described in write the information that lock comprises described Section Point after, also comprise:

Described spot lock management device receives the read lock that described file destination that described Section Point sends is corresponding or writes lock application request;

Described spot lock management device determine described first node application cross described file destination corresponding write lock, described file destination was carried out to up-to-date modification, described spot lock management device sends to described first node the lock of writing that described file destination is corresponding and reclaims order, to described first node reclaim described file destination corresponding write lock, described spot lock management device to described first node, reclaim described file destination corresponding write when lock, the data of the described file destination in the buffer memory of described first node are not write to dish and invalid;

Described spot lock management device sends read lock or writes lock authorization response to described Section Point, and described read lock or write the information that lock comprises described first node, so that described Section Point is according to the up-to-date global buffer table of the acquisition of information of described first node.

The buffer memory synchronous method of 13. 1 kinds of distributed type assemblies target file system, is characterized in that, comprising:

Described Section Point receives the global buffer table synchronization request message that described first node sends;

Described Section Point is according to described global buffer table synchronization request message, to described first node, return to global buffer table synchronous response message, in described global buffer table synchronous response message, comprise described up-to-date global buffer table, in described up-to-date global buffer table, preserve the up-to-date amendment record of described Section Point to described file destination.

14. methods according to claim 13, is characterized in that, described Section Point, according to described global buffer table synchronization request message, after described first node returns to global buffer table synchronous response message, also comprises:

Described Section Point receives the data synchronization request message that described first node sends, and described data synchronization request message is that described first node sends the up-to-date amendment record of described file destination according to Section Point described in described up-to-date global buffer table;

Described Section Point, according to described data synchronization request message, to described first node return data synchronous response message, comprises the Update Table of described Section Point to described file destination in described data sync response message.

15. 1 kinds of nodes, described node is first node, it is characterized in that, comprising:

Lock application module, for when client is carried out write access to the file destination on first node, to file destination described in spot lock management device application corresponding write lock, described in write lock and comprise the information of described file destination being carried out to the Section Point of up-to-date modification;

Cache table acquisition module, for the up-to-date global buffer table corresponding according to file destination described in the acquisition of information of described Section Point, preserves the up-to-date amendment record of described Section Point to described file destination in described up-to-date global buffer table;

File modification module, for the up-to-date amendment record of described file destination being modified to described file destination according to described Section Point, amended data are kept in the buffer memory of described first node, and described first node is updated to described up-to-date global buffer table to the amendment record of described file destination.

16. nodes according to claim 15, is characterized in that, described file modification module comprises judging unit, data capture unit and file modification unit;

Described judging unit, for determining whether to obtain the Update Table of described Section Point to described file destination according to described Section Point to the up-to-date amendment record of described file destination;

If, described acquiring unit obtains the Update Table of described Section Point to described file destination, the Update Table of the described file destination getting is stored in the buffer memory of described first node, the data of the described file destination stored on described first node and described Section Point are consistent, described file modification unit reads described data from the buffer memory of described first node, and described file destination is modified;

If not, described file modification unit reading out data from the buffer memory of described first node, modifies to described file destination.

17. nodes according to claim 16, it is characterized in that, described file destination comprises at least two data blocks, and described Section Point comprises the up-to-date amendment record of described file destination: described Section Point carried out the information of the data block of up-to-date modification to described file destination;

Described judging unit specifically for:

According to described Section Point, described file destination was carried out the information of the data block of up-to-date modification, determined that described first node carried out the data block of write access to described file destination and whether data block that described Section Point carried out up-to-date modification to described file destination is same data block;

If so, described data capture unit obtains the Update Table of described Section Point to described file destination.

18. nodes according to claim 15, is characterized in that, described cache table acquisition module specifically for:

To described Section Point, send global buffer table synchronization request message;

Receive the global buffer table synchronous response message that described Section Point sends, in described global buffer table synchronous response message, comprise described up-to-date global buffer table, in described up-to-date global buffer table, preserve the up-to-date amendment record of described Section Point to described file destination.

19. nodes according to claim 17, is characterized in that, described data capture unit specifically for:

According to the up-to-date amendment record of Section Point to described file destination described in described up-to-date global buffer table, to described Section Point, send data synchronization request message;

Receive the data sync response message that described Section Point sends, in described data sync response message, comprise the Update Table of described Section Point to described file destination.

20. nodes according to claim 15, is characterized in that, also comprise: file read module;

Described lock application module also for: when described first node carries out read access to described file destination, to read lock corresponding to file destination described in described spot lock management device application, in described read lock, comprise the information of described file destination being carried out to the described first node of up-to-date modification;

Described file read module, reads described file destination for the buffer memory from described first node.

21. nodes according to claim 15, is characterized in that, also comprise:

Receiver module, the global buffer table synchronization request message sending for receiving described Section Point, described global buffer table synchronization request message is that described Section Point sends to described first node need to carry out read access or write access to described file destination time;

Sending module, for according to described global buffer table synchronization request message, returns to global buffer table synchronous response message to described Section Point, comprises the up-to-date global buffer table after described renewal in described global buffer table synchronous response message.

22. nodes according to claim 21, is characterized in that, described receiver module also for:

Receive the data synchronization request that described Section Point sends, described data synchronization request message is that described Section Point sends the amendment record of described file destination according to first node described in the up-to-date global buffer table after described renewal;

Described sending module also for: to described Section Point, send data sync response, described data sync response comprises the Update Table of described first node to described file destination, so that described Section Point is kept at described first node in the buffer memory of described Section Point to the Update Table of described file destination, when carrying out read access or write access to described data, described Section Point can from the buffer memory of described Section Point, read described file destination.

23. nodes according to claim 15, is characterized in that, described lock application module specifically for:

To described spot lock management device, send the lock application request of writing that described file destination is corresponding, so that described spot lock management device is write lock described in described Section Point recovery;

What receive that described spot lock management device sends writes lock authorization response.

24. 1 kinds of spot lock management devices, is characterized in that, comprising:

Receiver module, the lock application request of writing corresponding to file destination sending for receiving described first node, described in to write lock application request be client when described file destination is carried out to write access, described first node sends to described spot lock management device;

Determination module, for determine Section Point application cross described file destination corresponding write lock, described Section Point carried out up-to-date modification to described file destination;

Sending module, for send the lock of writing that described file destination is corresponding to described Section Point, reclaim order, to described Section Point reclaim described file destination corresponding write lock, described spot lock management device to described Section Point, reclaim described file destination corresponding write when lock, the data of the described file destination in the buffer memory of described Section Point are not write to dish and invalid;

Described sending module also for: to described first node, send and write lock authorization response, described in write the information that lock comprises described Section Point.

25. spot lock management devices according to claim 24, is characterized in that, described receiver module also for:

Receive read lock application request corresponding to described file destination that described first node sends;

Described determination module also for: determine described first node application cross described file destination corresponding write lock, described first node carried out up-to-date modification to described file destination;

Described sending module also for: to described first node, send the lock of writing that described file destination is corresponding and reclaim order, to described first node reclaim described file destination corresponding write lock, described spot lock management device to described first node, reclaim described file destination corresponding write when lock, the data of the described file destination in the buffer memory of described first node are not write to dish and invalid;

Described sending module also for: to described first node, send read lock authorization response, described read lock comprises the information of described first node, so that described first node reads described file destination from the buffer memory of described first node.

26. spot lock management devices according to claim 24, is characterized in that, described receiver module also for:

Receive the read lock that described file destination that described Section Point sends is corresponding or write lock application request;

Described determination module also for: determine described first node application cross described file destination corresponding write lock, described first node carried out up-to-date modification to described file destination;

Described sending module also for: to described first node, send the lock of writing that described file destination is corresponding and reclaim order, to described first node reclaim described file destination corresponding write lock, described spot lock management device to described first node, reclaim described file destination corresponding write when lock, the data of the described file destination in the buffer memory of described first node are not write to dish and invalid;

Described sending module also for: to described Section Point, send read lock or write lock authorization response, described read lock or write the information that lock comprises described first node, so that described Section Point is according to the up-to-date global buffer table of the acquisition of information of described first node.

27. 1 kinds of nodes, described node is Section Point, it is characterized in that, comprising:

Receiver module, the global buffer table synchronization request message sending for receiving described first node;

Sending module, be used for according to described global buffer table synchronization request message, to described first node, return to global buffer table synchronous response message, in described global buffer table synchronous response message, comprise described up-to-date global buffer table, in described up-to-date global buffer table, preserve the up-to-date amendment record of described Section Point to described file destination.

28. nodes according to claim 27, is characterized in that, described receiver module also for:

Receive the data synchronization request message that described first node sends, described data synchronization request message is that described first node sends the up-to-date amendment record of described file destination according to Section Point described in described up-to-date global buffer table;

Described sending module also for: according to described data synchronization request message, to described first node return data synchronous response message, in described data sync response message, comprise the Update Table of described Section Point to described file destination.

Images