WO2015055035A1

WO2015055035A1 - Method and device for hashing metadata object

Info

Publication number: WO2015055035A1
Application number: PCT/CN2014/083044
Authority: WO
Inventors: 李永健; 高峰
Original assignee: 华为技术有限公司
Priority date: 2013-10-18
Filing date: 2014-07-25
Publication date: 2015-04-23
Also published as: CN103559224A

Abstract

Disclosed is a method for hashing a metadata object, comprising: determining whether or not a metadata object is a snapshot metadata object; if the metadata object is not a snapshot metadata object, then using a name of the metadata object as an input value in calculating a hash value of the metadata object; if the metadata object is a snapshot metadata object, then using a version name in the name of the metadata object as an input value in calculating a hash value of the metadata object; and, hashing the metadata object to a corresponding metadata server node on the basis of the hash value of the metadata object. Also disclosed is a device for hashing a metadata object. Provided are the advantages of reduced randomness of the metadata object being hashed to different metadata server nodes, increased access efficiency of snapshot data, and increased snapshot access performance and overall system performance.

Description

Method and device for hashing metadata objects The present application claims to be filed on October 18, 2013 with the Chinese Patent Office, application number 201310493842.0, and the invention name is "a method and device for hashing metadata objects" Priority of the Chinese Patent Application, the entire contents of which is incorporated herein by reference.

TECHNICAL FIELD The present invention relates to the field of communications technologies, and in particular, to a method and apparatus for hashing a metadata object.

BACKGROUND A snapshot is one of the important means for system disaster recovery backup. The snapshot is mainly used to implement online backup and recovery of data, and can quickly restore data to a specified time point when the system fails, and can allow users to access the snapshot. The snapshot version data (hereinafter referred to as snapshot data) of the metadata object generated by the protected metadata object at any point in time. In order to ensure the load balancing of each metadata server in the system and prevent a directory node from being too large or too hot, the current cluster system needs to split the directory metadata object to hash all metadata objects to different nodes. Since the snapshot is directory-based, if the directory node is split, the snapshot metadata object generated when the metadata object is protected by the snapshot is also hashed to a different metadata server node. In the prior art, a metadata object (including a normal metadata object and a snapshot metadata object) is hashed by the name of each metadata object (including a common element) in the same directory (for example, a directory whose node is too large). The name of the data object and the name of the snapshot metadata object are executed by HASH calculation, and all metadata objects are hashed to different nodes according to the result of the HASH calculation. In the prior art, since the names of the respective metadata objects in the same directory are different, the respective metadata objects are

HASH calculation results are also diverse, and each metadata object is hashed to different nodes with strong randomness. As shown in Figure 1, each metadata object in Figure 1 (including two general metadata objects Filel and File2 under the current metadata server node Dir and its corresponding snapshot metadata object Filel@snapl,

Filel@snap3, File2@snapl, File2@snap3, etc., where "Dir" Chinese can be defined as "directory", "File" Chinese can be defined as "file", and "snap" is an abbreviation of "snapshot". The Chinese can be defined as "snapshot". The "File" and "snap" in Figure 2 and Figure 4 below can also be in the same Chinese meaning. The names of the subsequent HASH calculations can be used to separate the metadata objects. Columns to different nodes, as shown in Figure 2, Dirl and Dir2 in Figure 2 are two different nodes. Since the names of the various metadata objects in Figure 1 are different, the HASH results of each metadata object are not the same, so that the graph All metadata objects in 1 are hashed to the Dirl or Dir2 nodes with strong randomness. In the prior art, different snapshot versions of the same metadata object are hashed to different nodes, the access cost of the snapshot data is large, the access performance is low, and the system performance is low.

SUMMARY OF THE INVENTION The technical problem to be solved by the embodiments of the present invention is to provide a metadata server node that hashes the metadata objects, and reduces the randomness of the metadata objects being hashed to different metadata server nodes, and the snapshot is improved. Data access efficiency, improved snapshot access performance and overall system performance.

A first aspect of the embodiments of the present invention provides a method for hashing a metadata object, including: determining whether a metadata object is a snapshot metadata object;

If the metadata object is not a snapshot metadata object, calculating a hash value of the metadata object by using the name of the metadata object as an input value;

If the metadata object is a snapshot metadata object, calculating a hash value of the metadata object by using a version name in a name of the metadata object as an input value;

The metadata object is hashed to a corresponding metadata server node based on a hash value of the metadata object.

With reference to the first aspect, in a first possible implementation, the snapshot metadata object is a snapshot version of a metadata object generated when the metadata object is protected by a snapshot;

The composition of the name of the snapshot metadata object includes: version name + delimiter + version number. With reference to the first possible implementation manner of the first aspect, in a second possible implementation, the determining whether the metadata object is a snapshot metadata object includes:

Determining whether the name of the metadata object includes a name identifier of the snapshot metadata object; if the determination result is yes, determining that the metadata object is a snapshot metadata object, and if the determination result is no, determining the element The data object is not a snapshot metadata object. With reference to the first aspect to any one of the second possible implementation manners of the first aspect, in a third possible implementation, the calculating the metadata object by using the name of the metadata object as an input value Hash values, including:

The name of the metadata object is used as an input value of the HASH calculation, and the metadata object is subjected to HASH calculation to determine a hash value of the metadata object.

With reference to any one of the first aspect to the second possible implementation manner of the second aspect, in a fourth possible implementation, the calculating, by using the version name in the name of the metadata object as the input value, The hash value of the metadata object, including:

The version name in the name of the metadata object is used as an input value of the HASH calculation, and the metadata object is subjected to HASH calculation to determine a hash value of the metadata object.

The second aspect of the embodiment of the present invention provides an apparatus for hashing a metadata object, including: a determining module, configured to determine whether the metadata object is a snapshot metadata object;

a first calculating module, configured to calculate a hash value of the metadata object by using a name of the metadata object as an input value when the determining result of the determining module is negative;

a second calculating module, configured to calculate a hash value of the metadata object by using a version name in a name of the metadata object as an input value when the determining result of the determining module is YES;

And a hashing module, configured to hash the metadata object to a corresponding metadata server node according to a hash value of the metadata object.

With reference to the second aspect, in a first possible implementation, the snapshot metadata object is a snapshot version of a metadata object generated when the metadata object is protected by a snapshot;

The composition of the name of the snapshot metadata object includes: version name + delimiter + version number. With reference to the first possible implementation manner of the second aspect, in a second possible implementation, the determining module is specifically configured to: when determining whether the metadata object is a snapshot metadata object,

Determining whether the name of the metadata object includes a name identifier of the snapshot metadata object, and if the determination result is yes, determining that the metadata object is a snapshot metadata object, and if the determination result is no, determining the element The data object is not a snapshot metadata object.

With reference to any one of the second aspect to the second possible implementation of the second aspect, in a third possible implementation, the first calculating module calculates the name of the metadata object as an input value When the hash value of a metadata object is used, specifically: The name of the metadata object is used as an input value of the HASH calculation, and the metadata object is subjected to HASH calculation to determine a hash value of the metadata object.

With reference to any one of the second aspect to the second possible implementation manner of the second aspect, in a fourth possible implementation, the second calculating module takes the version name in the name of the metadata object as an input When the value is calculated as the hash value of the metadata object, it is specifically used to:

The version name in the name of the metadata object is used as an input value of the HASH calculation, and the metadata object is subjected to HASH calculation to determine a hash value of the metadata object. The embodiment of the present invention may calculate a hash value of the metadata object according to the name of the metadata object and the version name of the snapshot metadata object, and hash the metadata object to the corresponding metadata server node according to the calculated hash value. The embodiment of the present invention may hash the current metadata object and its different versions of the snapshot metadata object to the same metadata server node, and reduce the randomness of the metadata object being hashed to different metadata server nodes. Improve the access efficiency of snapshot data, reduce the information interaction between the metadata server nodes due to snapshot access, improve the access performance of the snapshot and the overall performance of the system.

BRIEF DESCRIPTION OF THE DRAWINGS In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings to be used in the embodiments or the description of the prior art will be briefly described below, and obviously, in the following description The drawings are only some of the embodiments of the present invention, and those skilled in the art can obtain other drawings based on these drawings without any creative work.

1 is a schematic diagram of relationship between a snapshot protected metadata object and a generated snapshot metadata object in the prior art;

2 is a schematic diagram of a snapshot metadata object generated by a snapshot protected metadata object in a prior art being hashed to different nodes;

3 is a schematic flowchart of a method for hashing a metadata object according to an embodiment of the present invention; FIG. 4 is a metadata protected by a snapshot in an embodiment of a method for hashing a metadata object according to an embodiment of the present invention; FIG. 5 is a schematic structural diagram of an apparatus for hashing a metadata object according to an embodiment of the present invention; FIG. 6 is a schematic diagram of a terminal provided by an embodiment of the present invention; Example structure diagram. The technical solutions in the embodiments of the present invention are clearly and completely described in the following with reference to the accompanying drawings in the embodiments of the present invention. It is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. example. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present invention without departing from the inventive scope are the scope of the present invention.

Referring to FIG. 3, it is a schematic flowchart of an embodiment of a method for hashing a metadata object according to an embodiment of the present invention. The method for hashing a metadata object described in this embodiment includes the steps of:

S101. Determine whether the metadata object is a snapshot metadata object. If the determination result is no, step S102 is performed. If the determination result is yes, step S102 is performed.

S102. Calculate a hash value of the metadata object by using the name of the metadata object as an input value.

5103. Calculate a hash value of the metadata object by using a version name in a name of the metadata object as an input value.

5104. Hash the metadata object to a corresponding metadata server node according to a hash value of the metadata object.

In a specific implementation, the snapshot metadata object described in the embodiment of the present invention is a snapshot version of the metadata object generated when the metadata object is protected by the snapshot, and the snapshot version generated by one common metadata object may include multiple. The current main ways to implement snapshots include COW (Copy On Write) and ROW (Redirect On Write). If the method for implementing snapshots is COW, the data is protected by snapshots mainly referring to new data. When writing to a storage location for the first time, the data at the location is first read out and written to another location (the storage space reserved for the snapshot, which can be called the snapshot space, and the data stored in the space is Snapshot data), then write new data to the location; If the method of implementing the snapshot is ROW, the data is protected by snapshot mainly refers to the newly written data when the new data is first written to a storage location. Being redirected to a new storage location, the storage location of the original data becomes the snapshot space. The protection of the metadata object by the snapshot is also the same as the protection of the above data by the snapshot. In a specific implementation, the metadata object described in this embodiment may include: a file, a directory, and the like, as shown in FIG. 1 , where File1 and File2 are ordinary elements under the metadata server node Dir described in the embodiment of the present invention. Data object, Filel @snapl, Filel@snap3 in Figure 1 is the snapshot metadata object generated by the metadata object Filel, File2@snapl, File2@snap3 is the metadata pair. A snapshot metadata object generated by File2; Filel and File2 are two metadata objects under the metadata server node Dir, and Filel@snapl and Filel@snap3 are snapshot metadata objects of different versions of the same version generated by Filel. File2@snapl and File2@snap3 are snapshot metadata objects of the same version of the same version generated by File2. In a specific implementation, the namespace of the snapshot metadata object is unique to the system, and the naming rules of the names of the respective snapshot metadata objects are also uniquely determined. In some possible implementations, the naming rule for the name of the snapshot metadata object can be: version name

+ delimiter + version number, that is, the composition of the name of the snapshot metadata object includes: version name + delimiter + version number, where the delimiter can be "@". For example, in the snapshot metadata object Filel@snapl, Filel is the version name of the snapshot metadata object, @ is the delimiter, and snapl is the version number.

In some possible implementations, in order to cause all snapshot versions of the same file (or directory) under any metadata server node to be hashed to the same metadata server node, the user accesses the snapshot metadata object of the file. When you need to find these snapshot metadata objects on the same metadata server node, you do not need cross-node query. When hashing all snapshot metadata objects of the same file, you can use the same name as the metadata object hash. Enter a value, calculate the hash value of each metadata object, and hash each metadata object based on the hash value. In a specific implementation, when the metadata object is hashed, it may first determine whether the metadata object is a snapshot metadata object, and if it is a snapshot metadata object, calculate the name by using the version name in the name of the metadata object as an input value. The hash value of the metadata object, if it is not the snapshot metadata object, can determine that the metadata object is an ordinary metadata object, and then the name of the metadata object can be used as an input value to calculate the dispersion of the metadata object. Column value. Specifically, determining whether the metadata object is a snapshot metadata object may determine, according to the name of the metadata object, whether the name of the metadata object includes a name identifier of the snapshot metadata object, where the name identifier of the snapshot metadata object Specifically, it may be a delimiter in the name of the snapshot metadata object (for example, @ ). If it is determined that the name of the metadata object includes a delimiter in the name of the snapshot metadata object, the metadata object may be determined to be known. For the snapshot metadata object, if the name of the metadata object does not include the delimiter in the name of the object of the snapshot metadata, it can be determined that the metadata object is not the snapshot metadata object, but the current metadata object. . In a specific implementation, when the hash value of the metadata object is calculated by using the name of the current metadata object as an input value, specifically, the name of the metadata object is used as an input value calculated by HASH, and the metadata object is subjected to HASH calculation. Determine the hash value of the metadata object; when calculating the hash value of the snapshot metadata object, it will be faster The version name in the name of the metadata object is used as an input value of the HASH calculation, and the metadata object is subjected to HASH calculation to determine the hash value of the metadata object. For example, when hashing each metadata object in Figure 1, for the current metadata object Filel or File2, when calculating the hash value, you can directly calculate the hash of File1 or File2 by using Filel or File2 as the input value of HASH. For the snapshot metadata objects Filel@snapl and Filel@snap3, when calculating the hash value, the hash value of Filel@snapl and Filel@snap3 can be calculated by using the version name Filel in its name as the input value of HASH. Since the file values are used as HASH input values when calculating the hash values of Filel, Filel@snapl, and Filel@snap3, the calculated hash values of Filel, Filel@snapl, and Filel@snap3 are the same value. Similarly, for the metadata objects File2, File2@snapl, and File2@snap3 in Figure 1, since the hash values of File2, File2@snapl, and File2@snap3 are calculated as File2 as the input value of HASH, the calculated File2 is calculated. The hash values of File2@snapl and File2@snap3 are also the same value.

In some feasible implementations, after calculating the hash value of each metadata object, each metadata object may be hashed to the corresponding metadata service node according to the hash value of the metadata object. Specifically, the hash value of the same or similar hash value may be hashed to the same metadata server node by processing the hash value of each metadata object (for example, rounding), or the hash value may be different in size. Large metadata objects are hashed to different metadata server nodes, and so on. For example, when hashing each metadata object in Figure 1, since the hash values of the metadata objects Filel, Filel@snapl, and Filel@snap3 are the same value, the hash values of these metadata objects are used. When hashing, Filel, Filel@snapl, and Filel @snap3 are hashed to the same metadata server node; in addition, the hash values of the metadata objects File2, File2@snapl, and File2@snap3 are also the same value. Hash File2, File2@snapl, and File2@snap3 to the same metadata server node, as shown in Figure 4. Comparing Figure 2 with Figure 4, the metadata objects of all snapshot versions of the same file in Figure 2 may be hashed to different metadata server nodes, for example, the metadata version of Filel's snapshot version, Filel @snapl, and Filel @snap3 may be hashed to node Dirl and node Dir2 respectively. Users may need to query across snapshots to obtain complete snapshot data when accessing snapshot data. In Figure 4, all snapshot versions of the same file are in metamorphosis. Data objects are hashed to the same metadata server node. For example, the metadata objects Filel@snapl and Filel @snap3 of the snapshot version of Filel are hashed to the node Dirl, and the user accesses a file. All snapshot versions only need to be searched at the node of the file. Only the query on the node can obtain complete snapshot data without cross-nodes, which reduces the information interaction between the metadata server nodes due to snapshot access. Reduce the overhead of snapshot access.

The embodiment of the present invention mainly calculates the hash value of the metadata object by using the name of the metadata object and the version name of the snapshot metadata object as the input value of the HASH calculation, and hashing the metadata object according to the calculated hash value to The corresponding metadata server node. When the hash value of the metadata object and its different versions of the snapshot metadata object is calculated in the embodiment of the present invention, the same hash value can be obtained, and the current metadata object and its different versions of the snapshot element can be obtained. The data objects are hashed to the same metadata server node, which reduces the randomness of the metadata objects being hashed to different metadata server nodes, thereby improving the access efficiency of the snapshot data and reducing the snapshot access between the metadata server nodes. Increased information interaction, improved snapshot access performance and overall system performance.

FIG. 5 is a schematic structural diagram of an embodiment of an apparatus for hashing a metadata object according to an embodiment of the present invention. The device for hashing the metadata object described in this embodiment includes: a determining module 10, configured to determine whether the metadata object is a snapshot metadata object.

The first calculating module 20 is configured to calculate a hash value of the metadata object by using the name of the metadata object as an input value when the determining result of the determining module is negative.

The second calculating module 30 is configured to calculate a hash value of the metadata object by using a version name in the name of the metadata object as an input value when the determining result of the determining module is YES.

The hashing module 40 is configured to hash the metadata object to a corresponding metadata server node according to a hash value of the metadata object.

In some possible implementation manners, when the determining module determines whether the metadata object is a snapshot metadata object, the determining module is specifically configured to:

Determining, according to the name of the metadata object, whether the name of the metadata object includes a name identifier of the snapshot metadata object, and if the determination result is yes, determining that the metadata object is a snapshot metadata object, if the result is determined If not, it is determined that the metadata object is not a snapshot metadata object.

In some possible implementations, when the first computing module calculates the hash value of the metadata object by using the name of the metadata object as an input value, specifically:

The name of the metadata object is used as an input value of the HASH calculation, and the metadata object is subjected to HASH calculation to determine a hash value of the metadata object. In some possible implementation manners, when the second calculating module calculates the hash value of the metadata object by using the version name in the name of the metadata object as an input value, specifically:

In a specific implementation, the snapshot metadata object described in the embodiment of the present invention is a snapshot version of the metadata object generated when the metadata object is protected by the snapshot, where the snapshot version generated by one common metadata object may include multiple, Common metadata objects under the metadata server node described in the embodiments may include: files, directories, and the like. As shown in FIG. 1, Filel and File2 are ordinary metadata objects under the metadata server node Dir described in the embodiment of the present invention, and Filel @snapl and Filel@snap3 in FIG. 1 are generated by the metadata object Filel. The snapshot metadata object, File2@snapl, File2@snap3, is the snapshot metadata object generated by the metadata object File2; Filel and File2 are the two metadata objects under the metadata server node Dir, Filel@snapl and Filel@ Snap3 is a snapshot metadata object of the same version of the same version generated by Filel, according to the object. In a specific implementation, the namespace of the snapshot metadata object is unique to the system, and the naming rules of the names of the respective snapshot metadata objects are also uniquely determined. In some possible implementation manners, the naming rule of the name of the snapshot metadata object may be: a version name + a delimiter + a version number, that is, a composition of a name of the snapshot metadata object includes: a version name + a delimiter + a version number, where The delimiter can be "@". For example, in the snapshot metadata object Filel@snapl, Filel is the version name of the snapshot metadata object, @ is the delimiter, and snapl is the version number.

In some possible implementations, in order to cause all snapshot versions of the same file (or directory) under any metadata server node to be hashed to the same metadata server node, the user accesses the snapshot metadata object of the file. When you need to find these snapshot metadata objects on the same metadata server node, you do not need cross-node query. When hashing all snapshot metadata objects of the same file, you can use the same name as the metadata object hash. Enter a value, calculate the hash value of each metadata object, and hash each metadata object based on the hash value. In a specific implementation, when the metadata object is hashed, the determining module 10 may first determine whether the metadata object is a snapshot metadata object, and if it is a snapshot metadata object, the second computing module 30 is in the name of the metadata object. The version name as the input value, calculate the hash value of the metadata object, if not the snapshot metadata The object can determine that the metadata object is a normal metadata object, and the first computing module 20 can calculate the hash value of the metadata object by using the name of the metadata object as an input value. Specifically, the determining module 10 determines whether the metadata object is a snapshot metadata object, and determines, according to the name of the metadata object, whether the name of the metadata object includes a name identifier of the snapshot metadata object, where the snapshot metadata object is The name identifier may specifically be a delimiter in the name of the snapshot metadata object (for example, @ ). If it is determined that the name of the metadata object includes a delimiter in the name of the snapshot metadata object, it may be determined that the identifier is The metadata object is a snapshot metadata object. If the name of the metadata object does not include a delimiter in the name of the object of the snapshot metadata, it can be determined that the metadata object is not a snapshot metadata object, but the current Metadata object. In a specific implementation, when the first calculation module 20 calculates the hash value of the metadata object, the name of the current metadata object is used as the input value to calculate the hash value of the metadata object, specifically, the name of the metadata object is calculated as HASH. The input value, the HASH calculation is performed on the metadata object, and the hash value of the metadata object is determined; when the second calculation module 30 calculates the hash value of the snapshot metadata object, the version in the name of the snapshot metadata object may be The name is used as an input value of the HASH calculation, and the metadata object is subjected to HASH calculation to determine the hash value of the metadata object. For example, when hashing each metadata object in FIG. 1, for the current metadata object Filel or File2, when calculating the hash value, the first calculation module 20 can directly calculate Filel or File2 as the input value of HASH. The hash value of Filel or File2; and for the snapshot metadata objects Filel@snapl and Filel@snap3, when the hash value is calculated, the second calculation module 30 can use the version name Filel in the name of the snapshot metadata object as HASH. The input values calculate the hash values for Filel@snapl and Filel@snap3. Since the file values are used as HASH input values when calculating the hash values of Filel, Filel@snapl, and Filel@snap3, the calculated hash values of Filel, Filel@snapl, and Filel @snap3 are the same value. Similarly, for the metadata objects File2, File2@snapl, and File2@snap3 in Figure 1, since the hash values of File2, File2@snapl, and File2@snap3 are calculated as File2 as the input value of HASH, the calculated File2 is calculated. The hash values of File2@snapl and File2@snap3 are also the same value.

In some feasible implementation manners, after the hash value of each metadata object is calculated by the first computing module 20 or the second computing module 30, the hash module 40 may perform each element according to the hash value of the metadata object. The data object is hashed to the corresponding metadata service node. Specifically, the hash module 40 may treat the hash values of the respective metadata objects by processing (eg, rounding) the hash values to be the same or Nearly metadata objects are hashed to the same metadata server node, or metadata objects with large differences in hash values are hashed to different metadata server nodes. For example, when hashing each metadata object in Figure 1, since the hash values of the metadata objects Filel, Filel@snapl, and Filel @snap3 are the same value, based on the hash values of these metadata objects. Hash can be hashed to the same metadata server node by Filel, Filel@snapl, and Filel@snap3; in addition, the hash values of the metadata objects File2, File2@snapl, and File2@snap3 are the same value. Hash File2, File2@snapl, and File2@snap3 to the same metadata server node, as shown in Figure 4. Comparing Figure 2 with Figure 4, the metadata objects of all snapshot versions of the same file in Figure 2 may be hashed to different metadata server nodes, for example, the metadata version of Filel's snapshot version Filel@snapl and Filel@snap3 may be hashed to the node Dirl and the node Dir2 respectively. When the user accesses the snapshot data, it may need to query across the nodes to obtain the complete snapshot data. In Figure 4, all the snapshot versions of the same file are in the meta-version. Data objects are hashed to the same metadata server node. For example, the metadata objects Filel@snapl and Filel@snap3 of the snapshot version of Filel are hashed to the node Dirl. When the user accesses all snapshot versions of a file. It only needs to be searched in the node of the file, and only needs to query on the node to obtain complete snapshot data, without cross-node, reducing the information interaction between the metadata server nodes due to snapshot access, and reducing the snapshot access. s expenses.

The device for hashing the metadata object described in the embodiment of the present invention mainly calculates the hash value of the metadata object by using the name of the metadata object and the version name of the snapshot metadata object as the input value of the HASH calculation, and according to The computed hash value hashes the metadata object to the corresponding metadata server node. The device described in the embodiment of the present invention can obtain the same hash value when calculating the hash value of the metadata object and its different versions of the snapshot metadata object, and thus the current metadata object and its different versions can be obtained. The snapshot metadata object is hashed to the same metadata server node, which reduces the randomness of the metadata object being hashed to different metadata server nodes, thereby improving the access efficiency of the snapshot data and reducing the snapshot between the metadata server nodes. Increased information interaction for access, improving snapshot access performance and overall system performance.

FIG. 6 is a schematic structural diagram of an embodiment of a terminal according to an embodiment of the present invention. The terminal described in this embodiment includes:

The memory 100 is configured to store instructions. The processor 200 is configured to read an instruction from the memory, and perform an operation according to the instruction: determining whether the metadata object is a snapshot metadata object, and if the metadata object is not a snapshot metadata object, Calculating a hash value of the metadata object as an input value, and if the metadata object is a snapshot metadata object, calculating a version name in a name of the metadata object as an input value Deriving a hash value of the metadata object, hashing the metadata object to a corresponding metadata server node according to a hash value of the metadata object.

In some possible implementation manners, when determining, by the processor 200, whether the metadata object is a snapshot metadata object, the processor 200 is specifically configured to:

Determining, according to the name of the metadata object, whether the name of the metadata object includes a name identifier of the snapshot metadata object;

If the determination result is yes, it is determined that the metadata object is a snapshot metadata object, and if the determination result is no, it is determined that the metadata object is not a snapshot metadata object.

In some possible implementations, when the processor 200 calculates the hash value of the metadata object as the input value, the processor 200 is specifically configured to:

In some possible implementation manners, when the processor 200 calculates a hash value of the metadata object by using a version name in a name of the metadata object as an input value, specifically, the method is:

In a specific implementation, the snapshot metadata object described in the embodiment of the present invention is a snapshot version of the metadata object generated when the metadata object is protected by the snapshot, where the snapshot version generated by one common metadata object may include multiple, Common metadata objects under the metadata server node described in the embodiments may include: files, directories, and the like. As shown in FIG. 1, Filel and File2 are ordinary metadata objects under the metadata server node Dir described in the embodiment of the present invention, and Filel @snapl and Filel@snap3 in FIG. 1 are generated by the metadata object Filel. The snapshot metadata object, File2@snapl, File2@snap3, is the snapshot metadata object generated by the metadata object File2; Filel and File2 are the two metadata objects under the metadata server node Dir, Filel@snapl and Filel@ Snap3 is a snapshot metadata object of the same version with different version numbers generated by Filel. According to the object. In a specific implementation, the namespace of the snapshot metadata object is unique to the system, and the naming rules of the names of the respective snapshot metadata objects are also uniquely determined. In some possible implementation manners, the naming rule of the name of the snapshot metadata object may be: a version name + a delimiter + a version number, that is, a composition of a name of the snapshot metadata object includes: a version name + a delimiter + a version number, where The delimiter can be "@". For example, in the snapshot metadata object Filel@snapl, Filel is the version name of the snapshot metadata object, @ is the delimiter, and snapl is the version number.

In some possible implementations, in order to cause all snapshot versions of the same file (or directory) under any metadata server node to be hashed to the same metadata server node, the user accesses the snapshot metadata object of the file. Only need to find these snapshot metadata objects in the same metadata server node, without cross-node query, the processor 200 can hash all the snapshot metadata objects of the same file with the same name as the metadata object. The input value of the column calculation, the hash value of each metadata object is calculated, and each metadata object is hashed according to the hash value. In a specific implementation, when the processor 200 hashes the metadata object, it may first determine whether the metadata object is a snapshot metadata object, and if it is a snapshot metadata object, use the version name in the name of the metadata object as an input value. Calculating a hash value of the metadata object. If it is not a snapshot metadata object, determining that the metadata object is an ordinary metadata object, and then calculating the metadata by using the name of the metadata object as an input value. The hash value of the object. Specifically, the processor 200 determines whether the metadata object is a snapshot metadata object, and determines, according to the name of the metadata object, whether the name of the metadata object includes a name identifier of the snapshot metadata object, where the snapshot metadata object is The name identifier may specifically be a delimiter in the name of the snapshot metadata object (for example, @ ). If it is determined that the name of the metadata object includes a delimiter in the name of the snapshot metadata object, it may be determined that the identifier is The metadata object is a snapshot metadata object. If the name of the metadata object does not include a delimiter in the name of the object of the snapshot metadata, it may be determined that the metadata object is not a snapshot metadata object, but the current Metadata object. In a specific implementation, when the processor 200 calculates the hash value of the metadata object by using the name of the current metadata object as an input value, specifically, the name of the metadata object is used as an input value calculated by HASH, and the metadata object is performed on the metadata object. HASH calculation, determining a hash value of the metadata object; when the processor 200 calculates a hash value of the snapshot metadata object, the version name in the name of the snapshot metadata object may be used as an input value of the HASH calculation, and the metadata is The object performs HASH calculation, determining The hash value of this metadata object. For example, when hashing each metadata object in FIG. 1, for the current metadata object File1 or File2, when the processor 200 calculates its hash value, it can directly calculate Filel or File2 by using Filel or File2 as the input value of HASH. For the snapshot metadata objects Filel@snapl and Filel@snap3, when the processor 200 calculates its hash value, the version name Filel in its name can be used as the input value of HASH to calculate Filel@snapl and Filel@snap3. The hash value. Since the processor 200 calculates the hash values of Filel, Filel@snapl, and Filel@snap3 with Filel as the input value of HASH, the calculated hash values of Filel, Filel@snapl, and Filel@snap3 are the same value. . Similarly, for the metadata objects File2, File2@snapl, and File2@snap3 in FIG. 1, since the processor 200 calculates the hash values of File2, File2@snapl, and File2@snap3 using File2 as the input value of HASH, the calculation is performed. The obtained hash values of File2, File2@snapl, and File2@snap3 are also the same value.

In some feasible implementation manners, after the processor 200 calculates the hash value of each metadata object, each of the metadata objects may be hashed according to the hash value of the metadata object to the corresponding metadata service node. Specifically, the processor 200 may hash the metadata objects with the same or similar hash values to the same metadata server node by processing (eg, rounding) the hash values of the respective metadata objects, or hashing Metadata objects with large differences in value are hashed to different metadata server nodes, and so on. For example, when hashing each metadata object in Figure 1, since the hash values of the metadata objects Filel, Filel@snapl, and Filel @snap3 are the same value, the hash values of these metadata objects are used. When hashing, Filel, Filel@snapl, and Filel @snap3 are hashed to the same metadata server node; in addition, the hash values of the metadata objects File2, File2@snapl, and File2@snap3 are the same value. Hash File2, File2@snapl, and File2@snap3 to the same metadata server node, as shown in Figure 4. Comparing Figure 2 with Figure 4, the metadata objects of all snapshot versions of the same file in Figure 2 may be hashed to different metadata server nodes, for example, the metadata version of Filel's snapshot version Filel@snapl and Filel@snap3 may be hashed to the node Dirl and the node Dir2 respectively. When the user accesses the snapshot data, it may need to query across the nodes to obtain the complete snapshot data. In Figure 4, all the snapshot versions of the same file are in the meta-version. Data objects are hashed to the same metadata server node. For example, the metadata objects Filel @snapl and Filel@snap3 of the snapshot version of Filel are hashed to the node Dirl. When the user accesses all snapshot versions of a file. Just need to look up the node of the file, only need to be on that node By querying, the complete snapshot data can be obtained without cross-nodes, which reduces the information interaction between the metadata server nodes due to snapshot access, and reduces the overhead of snapshot access.

The terminal described in the embodiment of the present invention may calculate the hash value of the metadata object by using the name of the metadata object and the version name of the snapshot metadata object as the input value of the HASH calculation, and the metadata object according to the calculated hash value. Hash to the corresponding metadata server node. When the hash value of the metadata object and its different versions of the snapshot metadata object is calculated in the terminal described in the embodiment of the present invention, the same hash value can be obtained, and the current metadata object can be different from each other. The version of the snapshot metadata object is hashed to the same metadata server node, reducing the randomness of the metadata object being hashed to different metadata server nodes, thereby improving the access efficiency of the snapshot data and reducing the metadata server node Increased information interaction due to snapshot access, improving snapshot access performance and overall system performance.

A person skilled in the art can understand that all or part of the process of implementing the above embodiment method can be completed by a computer program to instruct related hardware, and the program can be stored in a computer readable storage medium, and the program is When executed, the flow of an embodiment of the methods as described above may be included. The storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM). The above is only the preferred embodiment of the present invention, and the scope of the present invention is not limited thereto, and thus equivalent changes made in the claims of the present invention are still within the scope of the present invention.

Claims

Rights request

1. A method for hashing metadata objects, characterized by: determining whether the metadata object is a snapshot metadata object;

If the metadata object is not a snapshot metadata object, use the name of the metadata object as the input value to calculate the hash value of the metadata object;

If the metadata object is a snapshot metadata object, use the version name in the name of the metadata object as an input value to calculate the hash value of the metadata object;

The metadata object is hashed to the corresponding metadata server node according to the hash value of the metadata object.

2. The method of claim 1, wherein the snapshot metadata object is a snapshot version of the metadata object generated when the metadata object is protected by a snapshot;

The name of the snapshot metadata object consists of: version name + delimiter + version number.

3. The method of claim 2, wherein the determining whether the metadata object is a snapshot metadata object includes:

Determine whether the name of the metadata object contains the name identifier of the snapshot metadata object;

If the judgment result is yes, it is judged that the metadata object is a snapshot metadata object; if the judgment result is no, it is judged that the metadata object is not a snapshot metadata object.

4. The method according to any one of claims 1 to 3, characterized in that, using the name of the metadata object as an input value to calculate the hash value of the metadata object includes:

The name of the metadata object is used as the input value of the HASH calculation, the HASH calculation is performed on the metadata object, and the hash value of the metadata object is determined.

5. The method according to any one of claims 1 to 3, characterized in that, using the version name in the name of the metadata object as an input value to calculate the hash value of the metadata object includes: The version name in the name of the metadata object is used as the input value of the HASH calculation, the HASH calculation is performed on the metadata object, and the hash value of the metadata object is determined.

6. A device for hashing metadata objects, characterized in that it includes: a judgment module, used to judge whether the metadata object is a snapshot metadata object;

The first calculation module is configured to use the name of the metadata object as an input value to calculate the hash value of the metadata object when the judgment result of the judgment module is no;

The second calculation module is configured to use the version name in the name of the metadata object as an input value to calculate the hash value of the metadata object when the judgment result of the judgment module is yes;

A hashing module, configured to hash the metadata object to the corresponding metadata server node according to the hash value of the metadata object.

7. The device of claim 6, wherein the snapshot metadata object is a snapshot version of the metadata object generated when the metadata object is snapshot protected;

8. The device according to claim 7, wherein the judgment module is specifically used to: when judging whether the metadata object is a snapshot metadata object:

Determine whether the name of the metadata object contains the name identifier of the snapshot metadata object. If the determination result is yes, then determine that the metadata object is a snapshot metadata object. If the determination result is no, then determine that the metadata object is a snapshot metadata object. The data object is not a snapshot metadata object.

9. The device according to any one of claims 6-8, wherein when the first calculation module uses the name of the metadata object as an input value to calculate the hash value of the metadata object, specifically: Used for:

10. The device according to any one of claims 6 to 8, characterized in that, the second calculation When the module uses the version name in the name of the metadata object as an input value to calculate the hash value of the metadata object, it is specifically used for:

The version name in the name of the metadata object is used as the input value of the HASH calculation, the HASH calculation is performed on the metadata object, and the hash value of the metadata object is determined.