Hitachi Solutions, Ltd.

Yasuhiro Kirihata

Koji Nakayama

Kouji Nakayama

Shimpei Nishida

Hideyuki Kashiwase

When an online storage service is used to expand a storage capacity of a file server, an amount of communication in synchronization processing and an amount of data retained on the online storage service are reduced to save an amount of charge. In a kernel module provided with a storage area on the online storage service, files are divided into block files and managed, and blocks overlapping with an already registered and saved block file group are not uploaded, but only configuration information of the files is changed. A mechanism is adopted, in which DBs for managing meta information and elimination of duplication are divided and managed, and only updated sections are appropriately uploaded.

Information management method, and computer for providing information

A file scan process is carried out, in parallel in a distributed manner, on a file server for which indices have been created once, to create a list of added, changed, or updated files. A file list of indexed scan target files is created by a distributed processing server. Directory URLs of the scan target files are acquired from the file list and analyzed to count a number of directories in each layer and determine a division unit for division of a scan process to be carried out on the scan target directories, based on a result of the counting.

Process of generating a list of files added, changed, or deleted of a file server

When an index is partitioned and managed in a plurality of indexes in a search system which deals with large amounts of data, the number of documents registered in each partitioned index may be varied, and inefficient data transfer may occur at the time of index reconfiguration when search nodes are added or deleted. In the present invention, mapping information is used to map search node IDs to slot regions obtained by equally partitioning a circular hash value space in which a start point and an end point of the hash value space are assumed to be identical, so that each document is mapped to each partitioned index.

Information processing system

When a storage capacity of a file server is expanded using an online storage service, elimination of an upper-limit constraint of the file size as a constraint of the online storage service and reduction in the communication cost are realized. A kernel module including logical volumes on the online storage service divides a file into block files at a fixed length and stores and manages the block files to prevent the upper-limit constraint of the file size. When a READ/WRITE request is generated for a mounted file system, only necessary block files are downloaded and used from the online storage service based on an offset value and size information to optimize the communication and realize the communication cost reduction.

File server apparatus, management method of storage system, and program

In a system which realizes to prevent leakage/loss of secret information by prohibiting a write operation to a secondary storage apparatus and a write operation to an external medium, an automatic collection of secret data to a server is executed, an existing application mode of PC is not damaged, and then an update of OS and an application is executed. The present invention places a secondary storage apparatus write control driver on the lower level than a file system, and redirects a write operation to the secondary storage apparatus, setting up a memory to be a primary cache, and cache data file on a cache server of a network destination to be a secondary cache. Thereby, the write operation to the secondary storage apparatus is not executed, and difference data is stored on the cache server, so that the automatic collection of secret data to the server can be realized.

Growing community of inventors

Tokyo, Japan

Yasuhiro Kirihata