CN1255748C - Metadata hierarchy management method and system of storage virtualization system - Google Patents

Metadata hierarchy management method and system of storage virtualization system Download PDF

Info

Publication number
CN1255748C
CN1255748C CN 200310111436 CN200310111436A CN1255748C CN 1255748 C CN1255748 C CN 1255748C CN 200310111436 CN200310111436 CN 200310111436 CN 200310111436 A CN200310111436 A CN 200310111436A CN 1255748 C CN1255748 C CN 1255748C
Authority
CN
Grant status
Grant
Patent type
Prior art keywords
directory
server
metadata
table
matching
Prior art date
Application number
CN 200310111436
Other languages
Chinese (zh)
Other versions
CN1545047A (en )
Inventor
金海�
贾永洁
吴松
冉龙波
王志平
周润松
Original Assignee
华中科技大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Grant date

Links

Abstract

本发明公开了一种存储虚拟化系统的元数据层次管理方法及其系统。 The present invention discloses a storage virtualization system metadata hierarchy management method and system. 该方法对元数据逻辑树进行层次管理,引入匹配表以减轻根目录服务器的压力,不仅可以快速准确的定向到用户所请求访问的目录,而且使系统达到更好的扩展性。 The method metadata tree hierarchy management logic, including a matching table to reduce the pressure on the root directory of the server, not only can quickly and accurately directed to the user access to the requested directory, but the system to achieve better scalability. 匹配表常驻内存,因为它只记录每个目录服务器的根入口地址,所以所占空间极小,而且管理方便。 Matching table permanent memory, because it records the root directory entry address of each server, so minimal space, and easy to manage. 由于匹配表所拥有的条目并不多,搜索匹配表对用户请求目录进行匹配的效率也很高。 Because of the efficiency of the matching table entry does not have much to search for a matching table to match the user requests a directory is also high. 其系统包括设置有匹配表及匹配表管理模块的元数据服务器和目录服务器;匹配表至少包括目录服务器名称和目录服务器上所保存的元数据子逻辑树的根目录。 Which is provided with a matching system comprising a matching table and a table management module and the directory server metadata server; matching table includes at least the name of the root directory and the directory server on the directory server stored metadata of the child logic tree. 本发明具有效率高、扩展性好和易于管理等特点。 The present invention has a high efficiency, scalability and ease of management features.

Description

一种存储虚拟化系统的元数据层次管理方法及其系统 A storage virtualization system metadata management method and system level

技术领域 FIELD

本发明属于计算机存储领域,具体涉及一种存储虚拟化系统的元数据层次管理方法及其系统。 The present invention belongs to the field of computer memory, particularly to a virtual storage system metadata hierarchy management method and system.

背景技术 Background technique

在现代高性能计算中,科学计算和军事应用对存储的要求越来越高,包括大容量、分布式、高性能和高可靠性。 In modern high-performance computing, scientific computing and military applications require more and more storage, including high-capacity, distributed, high performance and high reliability. 广域网的虚拟化存储管理方式将地理上分布的各种高性能存储系统集成为一体,形成庞大的分布存储空间,充分实现资源共享,提高资源利用率,有效解决存储数据的爆炸性增长和存储管理能力相对不足之间的矛盾。 Virtualized storage management of wide area network will be a variety of high-performance storage systems distributed geographically integrated together to form a large distributed storage space, full sharing of resources, improve resource utilization, effective solution to the explosive growth and storage management capabilities to store data the contradiction between the relative lack.

在一个广域存储系统中,这些海量的存储资源必须被有效的管理,从而引入元数据的概念。 In a storage system in a wide area, the mass storage resources must be efficiently managed, thereby introducing the concept of metadata. 元数据是描述数据的数据,它为系统提供对象物理位置与其逻辑名字之间的映射,一个逻辑文件可以对应多个物理文件副本。 Metadata is data that describes data, which provides a mapping between the object position and its logical name for the physical system, a logical file may correspond to multiple copies of the physical file. 此外,元数据还包括文件目录信息,文件信息,存储设备信息及相关的系统信息等等。 Further, the metadata further comprises a file directory information, file information, and information related to the storage device system information and the like.

在许多系统中比如SRB或者GridFTP,都采用层次目录结构管理元数据。 In many systems, such as the SRB or GridFTP, they are hierarchical directory management metadata structure. 当元数据变得庞大时,系统会同时启用多个元数据服务器,这样会导致以下问题的产生:一方面,逻辑树必须分布在这些元数据服务器上,而另一方面,目录服务器之间必须协同操作,因为要返回的元数据可能分布在不同的元数据服务器上,这样就对元数据的根目录服务器造成了压力。 When the metadata becomes large, the system will also enable multiple metadata servers, this will lead to the following problems: on the one hand, the logic tree must be distributed over the metadata server, on the other hand, must be between the directory server co-operation, to be returned because the metadata may be distributed over different metadata server, thus causing pressure on the root of the server metadata. 第三,如果根目录服务器出现故障,整个目录服务系统也就不能再正常运转了,很难做到系统的高可用。 Third, if the root server fails, the entire directory service system will no longer operate properly, it is difficult to achieve high availability of the system. 最后,这种结构使得元数据服务器的扩展也变得很困难。 Finally, this configuration makes extended metadata server is also difficult.

发明内容 SUMMARY

本发明的目的在于提供一种能克服上述缺陷的存储虚拟化系统的元数据层次管理方法。 Object of the present invention to provide a method for managing metadata hierarchy overcome the above disadvantages of storage virtualization system. 该方法具有更高的效率、更好的扩展性,并且更易于管理。 This method is more efficient, better scalability and easier to manage. 本发明还提供了实现该方法的系统。 The present invention also provides a system for implementing the method.

本发明提供的一种存储虚拟化系统的元数据层次管理方法,其步骤为:(1)把整个存储虚拟化系统划分为多个逻辑域,每个逻辑域内的元数据分布在多个目录服务器上;(2)在各逻辑域中设置一个元数据服务器,用于管理该域中的目录服务器;(3)在各元数据服务器中设置一个匹配表,该匹配表至少包括二个字段:目录服务器名称和目录服务器上所保存的元数据子逻辑树的根目录,用于直接将用户请求定向到其所在的目录服务器;(4)当用户提出请求时,从用户请求中提取要访问目录的路径;(5)在该域的匹配表中对路径进行长匹配,判断在匹配表中是否存在符合的目录服务器;(6)如果存在符合的目录服务器,直接访问匹配的目录服务器,并返回结果;否则显示出错信息。 A method for managing hierarchical metadata storage virtualization system according to the present invention provides, comprises the following steps: (1) the entire storage virtualization system is divided into a plurality of logical domains, the metadata for each logical domain distributed in a plurality of directory servers upper; (2) arranged in the logical domain a metadata server for managing the directory server in the domain; (3) provided in the table matches a metadata server, the matching table includes at least two fields: directory root name server and stored on the directory server metadata of the child logic tree for the user requests to direct it in the directory server; (4) when requested by the user, is extracted from the user requests access to the directory path; (5) on the path length matching in the matching table of the domain, whether or not directory server complies with the match table; (6) if the directory server complies with the presence of direct access to the matching directory server, and returns the result ; otherwise, it displays an error message.

实现上述方法的系统,被划分为多个逻辑域,每个逻辑域内的元数据分布在多个目录服务器上,其特征在于:每个逻辑域指定一台服务器作为元数据服务器,指定多台服务器作为保存元数据信息的目录服务器;元数据服务器内设置有匹配表及匹配表管理模块,匹配表至少包括目录服务器名称和目录服务器上所保存的元数据子逻辑树的根目录,匹配表管理模块用于匹配表中的记录进行维护、匹配表文件的解析以及根据用户请求的路径匹配到合适的目录服务器或显示出错信息。 System for implementing the method described above, is divided into a plurality of logical domains, the metadata for each logical domain is distributed over a plurality of directory servers, wherein: each of the specified logical domain server as a metadata server, multiple servers to saved as metadata information of the directory server; the metadata server is provided with a matching table and a matching table management module, a matching table includes at least the root directory on the directory server name and the directory server to save the metadata of the child logic tree, the matching table management module record for match table for maintenance, and interpret the match table file according to the matching user requests to the appropriate path of the directory server, or an error message.

本发明的元数据层次管理系统具有以下优点及效果。 Metadata level management system of the invention has the following advantages and effects.

(1)更高的效率:匹配表是常驻内存的,它保存的条目数受域内目录服务器数目所限,所以遍历匹配表对用户请求目录进行匹配的速度也很快。 (1) higher efficiency: matching table is a permanent memory, which stores the number of entries in the directory by the domain limited number of servers, so traverse the match table to user requests a directory for matching fast. 遍历后可以直接访问请求目录所在的目录服务器,而不用再途经根目录目录服务器,这将大大减轻根目录所在的目录服务器的负载,提高元数据的搜索的速度继而提高整个系统的效率。 After traversing the request can directly access the directory server directory is located, and no longer via the root directory of the server, which will greatly reduce the load on the root directory of the server is located, to improve the speed of searching metadata in turn improve the efficiency of the entire system.

(2)更好的扩展性:分层的树型管理模式使维护和管理易于扩展;当系统负载上升需要增加目录服务器时,只需要在匹配表中增加几个条目,标示出系统中新增加的目录服务器及其对应的根目录即可。 (2) better scalability: hierarchical management tree makes maintenance and management easy to expand; when the system load increases the need to increase the directory server, only need to add several entries in the Match Table, mark the new system increased the directory server and its corresponding root directory. 因为有了匹配表的调度,目录服务器可以达到相互独立,不需要感知域内其他目录服务器的存在。 Because of the match schedule, the directory server can be achieved independent of each other, you do not need to sense the presence of other directory servers in the domain. 这样,系统效率不会随目录服务器数目的增多而下降,达到系统效率的可扩展性。 Thus, the efficiency of the system will not change with an increase in the number of directory servers and decline to the scalability of the system efficiency.

(3)易于管理:匹配表的结构很简单,只记录了每个目录服务器的名字及存放在其上的元数据的根目录。 (3) ease of management: a matching table structure is very simple, only records the name of each directory server and stored in the root directory of metadata on it. 所以其添加删除修改都很方便。 So add delete modify easy.

当发明采用JAVA作为开发工具时,可以实现平台无关性;当存储虚拟化系统增加元数据副本时,由于副本分布情况能够完全在匹配表中显示出来,可以通过匹配表选择一个合适的目录服务器,从而有效的利用副本,提高系统效率。 When the invention is employed when JAVA as a development tool, can be realized platform independent; when the storage virtualization system increases copy of the metadata, since the copy of the distribution can be fully displayed in the matching table, you can select a suitable directory server by matching table, so as to effectively utilize a copy, improve system efficiency.

附图说明 BRIEF DESCRIPTION

图1为域内元数据逻辑树示例图;图2为本发明的元数据访问流程;图3为现有元数据逻辑树的访问方式;图4为本发明元数据逻辑树的访问方式;图5为元数据管理的结构图。 Figure 1 is a logic tree art example of FIG metadata; metadata access process of the present invention. FIG. 2; FIG. 3 is a conventional way to access the metadata of the logic tree; Fig metadata logic tree access method of the present invention; FIG. 5 It is a configuration diagram of a metadata management.

具体实施方式 detailed description

在我们的存储虚拟化系统GSP(G1obal Storage Provider)中,通过划分逻辑域对整个系统进行层次管理。 In our storage virtualization system GSP (G1obal Storage Provider), hierarchical management of the entire system by dividing the logical domains. 在每个域内,元数据分布在多个目录服务器上而,并按照类似Linux文件系统的组织方式形成一棵“本域元数据逻辑树”,图1是一个域元数据逻辑树示例。 In each domain, the metadata distributed over a plurality of directory servers, and the formation of a "gamut metadata present logical tree" according to a similar organization Linux file system, FIG. 1 is an exemplary logical tree gamut metadata. 图中,1.1,1.2,1.3,1.4表示域内的目录服务器;A0表示域内元数据逻辑树的根目录;A1,A2…,表示域内元数据逻辑树的子目录。 FIG, 1.1,1.2,1.3,1.4 represents the directory server in the art; A0 represents the root directory of metadata within the logic tree; A1, A2 ..., the metadata indicates that the subdirectory within the logic tree.

每个用户都隶属于一个域,用户的根目录绑定到域内一个子目录上,用户登陆时,系统自动定向到用户的根目录去。 Each user belongs to a domain, the root user bound to a subdirectory of the domain, user login, the system automatically directed to the user's root directory to.

建立基于目录树结构的元数据目录服务器(Directory Server)来存储元数据及其副本。 The establishment of a directory tree structure based on metadata Directory Server (Directory Server) to store metadata and copy. 但是在广域网的范围内,随着系统中资源的增长,文件和目录的信息会变得越来越庞大,很显然,目录服务器组织得不好,就会成为系统的瓶颈。 However, within the scope of the WAN, with the growth of system resources, file and directory information will become increasingly large, it is clear that the directory server poorly organized, it will become the bottleneck of the system. 所以如何快速准确的定向到用户的根目录就是一个十分突出的问题,这将影响到元数据的搜索速度进而影响整个系统的效率。 So how quickly and accurately directed to the user's home directory is a very prominent issue, which will affect the speed of the search metadata thereby affecting the efficiency of the entire system. 在我们的系统中,引入匹配表来解决上述问题。 In our system, the introduction of a matching table to solve the problem.

在每个域中设置一个元数据服务器管理域中的多个目录服务器,匹配表保存在元数据服务器上。 Set metadata management server in each domain a domain multiple directory servers, on the matching table stored in the metadata server. 匹配表是一个常驻内存的数据结构,记录了每个目录服务器的名字及存放在其上的元数据的根目录;如果一个目录服务器上面保存了某个目录的副本,也把这个目录放到匹配表中去。 Matching table is a memory-resident data structures, record the name of each directory server and stored in the root directory on the metadata; save a copy of a directory if a directory server above, but also to put this directory match table to go. 在匹配表中,还可以设置索引字段,作为某个条目在匹配表中的唯一标识。 In the matching table, the index field may be provided, as an entry uniquely identifies a match table.

当系统收到用户的请求时,首先在元数据服务器上遍历匹配表,根据用户所请求访问的路径对匹配表包含的所有路径做一个最长匹配,找到用户要访问的目录所在的目录服务器,直接访问这个目录服务器,而不用通过域内元数据根目录所在的目录服务器依次查找;如果元数据存在副本,还可以从匹配表中选择一个合适的目录服务器给用户访问。 When the system receives the user's request, the metadata server to traverse on the match table, according to the path requested by the user access to all the paths to make a matching table contains the longest match, find the directory server directory users want to access is located, direct access to the directory server, rather than sequentially directory to find the metadata server domain root directory is located; if there is a copy of the metadata, you can also choose an appropriate directory server from the match table for user access. 元数据访问流程如图2所不。 Metadata Access Scheme 2 are not shown in FIG.

图1列举了一个逻辑树的例子。 Figure 1 illustrates an example of a logic tree. 这个逻辑树代表一个域内的文件目录结构,分布在四个目录服务器上。 File directory structure of the tree represents a logical domain, distributed over four directory server. 其中,逻辑树的根/A0存放在目录服务器1.1上,目录/A0/B1存放在目录服务器1.2上,目录/A0/A1/A2/C11和目录/A0/A1/A2/C12放在目录服务器1.3上,目录/A0/B1/B2/B3/B4/D1放在目录服务器1.4上。 Wherein the logic tree root / A0 1.1 stored on the directory server, the directory / A0 / B1 1.2 stored on the directory server, the directory / A0 / A1 / A2 / C11 and directory / A0 / A1 / A2 / C12 in the directory server 1.3, the directory / A0 / B1 / B2 / B3 / B4 / D1 1.4 on the directory server. 表2显示了与图1逻辑树对应的匹配表。 Table 2 shows the matching table and a logical tree corresponding to FIG.

假设用户想访问/A0/B1/B2/B3/B4/D1/D2,如果按照现存的元数据访问方式,首先会从域内逻辑树的根目录所在的目录服务器1.1开始查找,历经目录服务器1.3,到达目录服务器1.4,然后再把结果返回给用户,如图3所示。 Assume that the user wants to access / A0 / B1 / B2 / B3 / B4 / D1 / D2, if in accordance with the existing metadata access, will first start looking from Directory Server 1.1 directory logic tree domain where, after Directory Server 1.3, reach 1.4 directory server, and then sends the results back to the user, as shown in FIG. 图中,R表示用户请求访问的文件/目录的路径;1.1,1.2,1.3,1.4表示本域中的目录服务器;A0表示域内元数据逻辑树的根目录;B1,D1…,表示域内元数据逻辑树中的目录;虚线箭头和r1,r2…表示用户访问步骤。 FIG, R represents a user requesting access path of the file / directory; 1.1,1.2,1.3,1.4 this domain represents the directory server; A0 represents the root directory of metadata within the logic tree; B1, D1 ..., represents metadata art logical directory tree; and dashed arrows r1, r2 ... represents a step of accessing a user. 但是如果采取本发明所用的匹配表方式,系统收到用户的请求后,将首先查找匹配表,根据用户请求的路径做一个最长匹配,这样就可以得到元数据在目录服务器1.4上面。 However, if you take the matching table used in the embodiment of the present invention, the system receives the user's request, the first match lookup tables, according to the path user's request to do a longest match so that the metadata can be obtained above 1.4 in the directory server. 系统将直接访问目录服务器1.4,而不需要再通过根目录入口1.1,如图4所示。 The system directly access the directory server 1.4, without the need to 1.1 by the root directory entry, as shown in FIG. 图中,1表示域内的所有的目录服务器,2表示域内元数据服务器GNS,3表示匹配表,虚线箭头和r1,r2表示用户访问步骤。 FIG, 1 shows all of the directory server domain, region 2 represents the metadata server GNS, 3 represents a matching table, and dashed arrows r1, r2 represents a user access step.

为了达到有效搜索的目的,所有目录服务器都必须自己保存域内元数据树的逻辑结构。 For the purpose of effective search of all directory servers must save yourself the logical domain of metadata tree. 例如,目录服务器1.4仍然会保存域内元数据树的根,但它只是一个路径,并不保存根的内容,这样当一个搜索请求到达1.4的时候就不需要对请求做任何变换。 For example, Directory Server 1.4 will still save the metadata tree root domain, but it is only a path, and do not save the contents of the root, so that when a search request reaches 1.4 do not need to do any conversion request. 这个根条目由于内容为空,只需要占用极少的存储空间,并且维护起来也很方便。 The root entry because the content is empty, only takes up little storage space, and maintenance is also very convenient.

匹配表的内容存放在一个文件中,当整个系统启动的时候,把文件中保存的匹配表读入内存。 Matching table of contents stored in a file, when the whole system starts to save the file matches the table read into memory. 匹配表的结构和功能决定了一般情况下都是在内存中对匹配表进行读操作,这就不会涉及到对匹配表文件的读写。 Matching table structure and function determines the read operation is generally matching table in memory, which is not related to the read and write files match table. 只有在新增目录服务器或者现有目录服务器的根目录发生变化的情况下才对匹配表进行修改,进而对文件进行修改,并不需要频繁读写匹配表的文件,大大节省了访问匹配表的时间。 Only when the root of the new directory server or an existing directory server changes fishes matching table is modified, and then modify the file, it does not require frequent reading and writing files matching table, saving a match access list time.

在元数据服务器中添加一个匹配表管理模块,对匹配表中的记录进行维护,具体操作包括记录的添加、删除、更新操作,匹配表文件的解析以及根据用户请求的路径匹配到合适的目录服务器等等,具体解释如下。 Adding metadata server in a matching table management module for matching records in the table maintenance, specific operations add including records, delete, update, interpret the match table file and matched to the appropriate directory server according to a path requested by the user etc., explained in detail as follows.

匹配表记录的添加:当域内新增一个目录服务器的时候,向匹配表中添加一条记录;如果某个目录服务器中的一个子目录过大,需要迁移到另外一个目录服务器去,也要通知匹配表,增加一条相关记录,以记录这个子目录和其迁移到的目录服务器。 Add a matching table records: When a new directory server in the domain when adding a record to match the table; if a directory server in a subdirectory is too large, need to move to another directory server, and shall notify match table, add a related record to record this subdirectory and their migration to the directory server.

匹配表记录的删除:当某个目录服务器不再服务于某个域时,从匹配表上删除相应的记录;匹配表记录的修改:当某个目录服务器上保存的目录树根目录发生变化时,修改相应的记录;匹配表文件的解析:当系统启动的时候,把匹配表文件中的内容读到字符串数组中,常驻内存;匹配用户请求的路径:在匹配表中查找用户请求访问的路径所在的目录服务器。 Delete table records match: When a directory server is no longer serving in a domain, delete the record from the matching table; amend matching table records: When stored on a directory server directory tree root changes modify the respective record; match table file parsing: when the system starts, the contents of the match table file to read a string array, resident memory; the user requests a path matching: Finding the user requests access match table the path to the directory server is located.

使用集群系统中的5个节点构建一个元数据层次管理系统,其基本配置如表1所示。 Use the 5 cluster system to build a metadata node-level management system, its basic configuration as shown in Table 1.

表1 各节点的硬件及网络配置其中,一台作为元数据服务器,另外4个作为目录服务器。 Table 1 each node hardware and network configuration wherein, as a metadata server, another 4 as a directory server. 元数据服务器负责匹配表的维护,包括生成添加删除更新等一系列的操作。 The metadata server is responsible for maintaining match tables, including adding generating a series of operations to remove the update and so on. 目录服务器负责保存元数据信息。 Directory Server is responsible for storing metadata information.

具体实施如下:其中一个节点充当元数据服务器,装载保存匹配表的文件和匹配表管理模块;其余四个节点充当1.1,1.2,1.3,1.4,装载LDAP数据库,保存域内的元数据。 Specific embodiments are as follows: a node which acts as a metadata server, load and save files matching table the matching table management module; the other four nodes serve as 1.1,1.2,1.3,1.4 loading LDAP database, metadata stored domain.

依照图5,4.1表示一个逻辑域Domain,其后的阴影部分4.2和4.3代表和4.1结构相同的多个逻辑域;5表示匹配表管理模块,虚线箭头表示元数据服务器和目录服务器之间的互访。 In accordance with FIG. 5,4.1 represents a logical domain Domain, followed by 4.2 and 4.3 shaded representation and a plurality of logical domains same structure 4.1; 5 represents the matching table management module, dashed arrows indicate the interaction between the metadata server, and the directory server visit.

整个系统的配置说明如下:(1)匹配表包括两个字段,其示例如表2。 Description of the system configuration as follows: (1) the matching table includes two fields, examples of which are as shown in Table 2.

表2 匹配表示例各字段解释如下:路径:目录服务器保存的元数据子逻辑树的根目录路径;目录服务器名称:此目录服务器的名字,域内各目录服务器名称唯一。 Table 2 shows an example of matching of each field is as follows: Path: Root directory path stored in the directory server metadata of the child logic tree; directory server name: Name of the directory server, the server names are unique within each directory.

(2)域内元数据保存在四个目录服务器上,具体分布示例如表3。 (2) within four metadata in the directory on the server, the distribution of the specific examples shown in Table 3.

表3 匹配表示例 Table 3 shows an example of matching

(3)匹配表运行示例系统启动时解析保存匹配表的文件,读至以下三个字符串数组,常驻内存:Sindex[i]:保存匹配表中第i个记录的索引值,匹配表中的索引值递增。 (3) matching table to run the sample analytical system holds the matching table at startup file is read to the following three array of strings, the permanent memory: Sindex [i]: matching table stored in recording an i-th index value, match table the index value is incremented. 它并非必需,但是可以简化和方便程序的编写;Spath[i]:保存第i个条目中字段“路径”的内容;sDS[i]:保存第i个条目中字段“目录服务器名称”的内容。 It is not necessary, but can be simplified and the preparation of convenient procedures; Spath [i]: save the contents of the field "path" i-th entry; sDS [i]: Save the i-th entry fields "directory server name" contents .

Sindex[i],spath[i],sDS[i]是相互对应的一套。 Sindex [i], spath [i], sDS [it] is set to correspond to each other. 从数组[1]开始记录文件。 From the array [1] to start recording files. sindex[0]的值为表中存在的条目数;spath[0]为空;sDS[0]为空。 Number sindex [0] value is present in the table entry; spath [0] is empty; sDS [0] is empty.

收到用户请求访问的文件或目录路径path后,首先遍历数组spath,如果spath[i]的值为path,返回sDS[i];如果spath中不存在path,则返回path的父目录所在的目录服务器们,或者父父目录直至整个域的根目录。 Upon receipt of a file or directory path to the user requesting access path, first through the array SPATH, if spath [i] is the path, return sDS [i]; if SPATH path does not exist, the directory of the parent directory path where the return they server, or parent parent directory until the root of the entire domain.

Claims (4)

  1. 1.一种存储虚拟化系统的元数据层次管理方法,其步骤为:(1)把整个存储虚拟化系统划分为多个逻辑域,每个逻辑域内的元数据分布在多个目录服务器上;(2)在各逻辑域中设置一个元数据服务器,用于管理该域中的目录服务器;(3)在各元数据服务器中设置一个匹配表,该匹配表至少包括二个字段:目录服务器名称和目录服务器上所保存的元数据子逻辑树的根目录,用于直接将用户请求定向到其所在的目录服务器;(4)当用户提出请求时,从用户请求中提取要访问目录的路径;(5)在该域的匹配表中对路径进行长匹配,判断在匹配表中是否存在符合的目录服务器;(6)如果存在符合的目录服务器,直接访问匹配的目录服务器,并返回结果;否则显示出错信息。 Metadata A storage management method of hierarchical virtualization system, the steps of: (1) the entire storage virtualization system is divided into a plurality of logical domains, each logical metadata art distributed across multiple directory servers; (2) disposed in the logical domain a metadata server for managing the directory server in the domain; (3) provided in the table matches a metadata server, the matching table includes at least two fields: the name of the directory server root directory server and stored on the metadata of the child logic tree for the user requests to direct it in the directory server; (4) when requested by the user, extracts the directory path to be accessed from the user request; (5) of the path in the matching table of the domain length matching, whether or not directory server complies with the match table; (6) if the directory server complies with the presence of direct access to the matching directory server, and returns the result; otherwise error message is displayed.
  2. 2.根据权利要求1所述的方法,其特征在于:所述匹配表中还设置有目录副本。 2. The method according to claim 1, wherein: the matching table is also provided with a copy of the directory.
  3. 3.根据权利要求1所述的方法,其特征在于:所述匹配表中还设置有索引字段。 3. The method according to claim 1, wherein: the table is also provided with a matching index field.
  4. 4.一种存储虚拟化系统的元数据层次管理系统,该系统被划分为多个逻辑域,每个逻辑域内的元数据分布在多个目录服务器上,其特征在于:每个逻辑域指定一台服务器作为元数据服务器(2),指定多台服务器作为保存元数据信息的目录服务器;元数据服务器(2)内设置有匹配表(3)及匹配表管理模块(5),匹配表(3)至少包括目录服务器名称和目录服务器上所保存的元数据子逻辑树的根目录,匹配表管理模块(5)用于匹配表(3)中的记录进行维护、匹配表文件的解析以及根据用户请求的路径匹配到合适的目录服务器或显示出错信息。 A storage virtualization system metadata hierarchy management system is divided into a plurality of logical domains, the metadata for each logical domain is distributed over a plurality of directory servers, wherein: each logical field specifies a server as meta data server (2), specify several servers as stored metadata information of the directory server; metadata server is provided with a matching table (3) and the matching table management module (5) (2), matching table (3 ) comprising at least a root directory on a server name and the directory server stored metadata of the child logic tree, the matching table management module (5) for matching table record (3) for maintenance, and interpret the match table file according to a user matching route request to the appropriate directory server, or an error message.
CN 200310111436 2003-11-24 2003-11-24 Metadata hierarchy management method and system of storage virtualization system CN1255748C (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN 200310111436 CN1255748C (en) 2003-11-24 2003-11-24 Metadata hierarchy management method and system of storage virtualization system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN 200310111436 CN1255748C (en) 2003-11-24 2003-11-24 Metadata hierarchy management method and system of storage virtualization system

Publications (2)

Publication Number Publication Date
CN1545047A true CN1545047A (en) 2004-11-10
CN1255748C true CN1255748C (en) 2006-05-10

Family

ID=34336101

Family Applications (1)

Application Number Title Priority Date Filing Date
CN 200310111436 CN1255748C (en) 2003-11-24 2003-11-24 Metadata hierarchy management method and system of storage virtualization system

Country Status (1)

Country Link
CN (1) CN1255748C (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101826054B (en) 2009-03-04 2011-12-07 安凯(广州)微电子技术有限公司 Micro memory Memory Management System

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4808409B2 (en) 2005-01-14 2011-11-02 株式会社日立製作所 A sensor network system, sensor data search method, and program
KR100610904B1 (en) 2005-03-03 2006-08-02 엘지전자 주식회사 Meta data parsing method for providing multimedia service and handset using thereof
JP2008078713A (en) 2006-09-19 2008-04-03 Sony Corp Recording apparatus and method, program, and reproducing apparatus and method
CN100595761C (en) 2007-12-29 2010-03-24 中国科学院计算技术研究所 Metadata management method for splitting name space
US8423581B2 (en) 2009-10-07 2013-04-16 International Business Machines Corporation Proxy support for special subtree entries in a directory information tree using attribute rules
CN101697168B (en) 2009-10-22 2011-10-19 中国科学技术大学 Method and system for dynamically managing metadata of distributed file system
CN102024016B (en) * 2010-11-04 2013-03-13 曙光信息产业股份有限公司 Rapid data restoration method for distributed file system (DFS)
CN102982151B (en) * 2012-11-27 2015-04-01 南开大学 Method for merging multiple physical files into one logic file
WO2014101000A1 (en) * 2012-12-26 2014-07-03 华为技术有限公司 Metadata management method and system
CN103870588B (en) * 2014-03-27 2016-08-31 杭州朗和科技有限公司 A method for use in a database and means

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101826054B (en) 2009-03-04 2011-12-07 安凯(广州)微电子技术有限公司 Micro memory Memory Management System

Also Published As

Publication number Publication date Type
CN1545047A (en) 2004-11-10 application

Similar Documents

Publication Publication Date Title
Padhy et al. RDBMS to NoSQL: reviewing some next-generation non-relational database’s
US5412808A (en) System for parsing extended file names in an operating system
US7136867B1 (en) Metadata format for hierarchical data storage on a raw storage device
US6678705B1 (en) System for archiving electronic documents using messaging groupware
US5483652A (en) Mechanism for locating without search discrete application resources known by common name only in a distributed network computing environment
US6782389B1 (en) Distributing files across multiple, permissibly heterogeneous, storage devices
US6915302B1 (en) Method, system, and program for accessing files in a file system
US4611272A (en) Key-accessed file organization
US6647393B1 (en) Dynamic directory service
US7350192B2 (en) Method and system to automatically generate software code
US20080005145A1 (en) Data processing
US20020065810A1 (en) File system translators and methods for implementing the same
US7546319B1 (en) File system consistency checking in a distributed segmented file system
US20050049994A1 (en) Systems and methods for the implementation of a base schema for organizing units of information manageable by a hardware/software interface system
US5617568A (en) System and method for supporting file attributes on a distributed file system without native support therefor
US6647391B1 (en) System, method and article of manufacture for fast mapping from a propertied document management system to a relational database
US20080256090A1 (en) Dynamic directory service
US20040059866A1 (en) System and method for representing named data streams within an on-disk structure of a file system
US20040167880A1 (en) System and method for searching a virtual repository content
US6286010B1 (en) Methods and apparatuses for interaction between schemata
US6026474A (en) Shared client-side web caching using globally addressable memory
US8261020B2 (en) Cache enumeration and indexing
US7328225B1 (en) System, method and computer program product for multi-level file-sharing by concurrent users
US6366921B1 (en) System and method for data manipulation in a dynamic object-based format
US20030191745A1 (en) Delegation of metadata management in a storage system by leasing of free file system blocks and i-nodes from a file system owner

Legal Events

Date Code Title Description
C06 Publication
C10 Request of examination as to substance
C14 Granted
C19 Lapse of patent right due to non-payment of the annual fee