Hitachi, Ltd.

Morihiro Iwata

Masashi Tsuchida

Yukio Nakano

Norihiro Hara

Yoshito Kamegi

Masato Asami

Susumu Kobayashi

Nobuyuki Yamashita

Nobuo Beniyama

Eiji Kaneko

Masataka Kan

Kentaro Chigusa

Kenichi Endo

Kenichi Chadani

Yuuji Hirose

The present invention grasps in advance a portion in which traffic is concentrated and which may become a bottleneck and simplifies management of an access path. In a storage network system that forms multiple paths between a large-capacity data storage device and a host computer that employs the data storage device, and transfers data by means of a selected path, access paths are managed by a management server. The management server selects a second access path that can be used as a replacement of a first access path that connects the data storage device with the host computer, and calculates the quantity of traffic on the first access path which is associated with a device disposed on the first access path. Also, the management server calculates a variation of the quantity of traffic on the second access path which is associated with a device disposed on the second access path taking the calculated quantity of traffic into consideration. Then, the management server determines whether the device on the second access path causes a bottleneck or not, according to the calculated value of the variation of traffic on the second access path. The management servertransfers the determined result to the outside, for example, a display device, and notifies a manager of the determined result.

Method and an apparatus for managing access path

The amount of work processing (main processing amount) of a database is acquired and accumulated. A value (main processing index) indicating the transition of the amount of main processing amount available in a time frame that is presumed to be required for backup is calculated by using the amounts of main processing accumulated during the latest period of time. The main processing index is compared with the backup executability reference value which is a threshold value determining the backup executability. If the main processing index is smaller than the threshold value, backup is executed. If the main processing index is larger than the threshold value, backup is not executed but the amount of main processing is again acquired to calculate the main processing index, and the main processing index is compared with the backup executability reference value.

Method for determining execution of backup on a database

A parallel database system retrieval method and system for transferring from database operation servers to a front end server only those data items that are used in subsequent processing, thereby shortening the query time when the unused data are massive. The invention allows an initial retrieval query to transfer only location information about the target data from a database operation server to the front end server. Using the location information the invention permits a subsequent query to get variable-bound location information so that the latter query will retrieve a desired sub-data item according to the acquired location information and dictionary information about the sub-data item locations within the data.

Querying database system to execute stored procedures using abstract data type attributes, retrieving location information of data, sub-data between first and second servers

In display processing and input processing of multidimensional aggregate patterns, a graph including coordinate axes for displaying aggregation levels of respective dimensions is used. An aggregate pattern of each aggregate cube is displayed as a figure on the graph. By using position relations on the graph, the relation between pre-aggregate data and aggregated data is visualized. On the basis of the position relations on the graph, the user can select a pre-aggregate cube.

Method and system for displaying multidimensional aggregate patterns in a database system

A database processing method for defining and operating data of file reference type which references a file. A column data insert request for specifying position information indicative of a storage location of a file and watch information for watching the file is received from a database client. A request for watching a file indicated by the position information is issued in accordance with the watch information. The position information and the watch information are stored in a database. The foregoing operations permit data in an external file placed outside a database to be treated as data equivalent to ordinary column data stored in the database to improve the integrity of the database.

Database processing method, apparatus for implementing same, and medium containing processing program therefor

A request accepting server accepts a processing request of data of a predetermined type, analyzes the request, and allocates a process corresponding to the request to either one of the request accepting server and at least one database processing server. A handle is a control block to manage a storage area of data as a processing target and has a server in which the processing target data has been stored and information regarding its storing position. When the processing target data corresponding to the processing request is processed in accordance with the processing request, the request accepting server executes a process according to the processing request by using a handle regarding the processing target data. When an entity of the processing target data is necessary the entity of the processing target data is obtained on the basis of the handle and is processed.

Database processing method and apparatus using handle

Querying parallel database system that execute stored procedures using abstract data type attributes, retrieving location information of data, sub-data between first and second server

An overloaded routines resolving method of analyzing an invocation for the group of overloaded routines to determine the routine to be called, the method including: the routine group sorting step of sorting the group of routines on the basis of the predetermined precedence; the reference utilization pruning step of determining the routine which has the highest precedence of a data type of an argument from among the routines which can be called for the types in the definition of the arguments of the invocation from the sorted routines and of deleting any of the routines each of which has the lower precedence than that of the routine thus determined from the candidates; and the sort order characteristic pruning step of comparing the data types of the arguments among the routines which still remain after having executed the reference utilization pruning step to delete any of the routines each of which has no possibility of being called from the candidates. From the foregoing, the necessary minimum candidate can be selected efficiently in the overloaded routines resolution which is carried out when analyzing the query for the database.

Method of resolving overloaded routines, system for implementing the same and medium for storing processing program therefor

A request accepting server accepts a processing request of data of a predetermined type, analyzes the request, and allocates a process corresponding to the request to either one of the request accepting server and at least one database processing server. A handle is a control block to manage a storage area of data as a processing target and has a server in which the processing target data has been stored and information regarding its storing position. When the processing target data corresponding to the processing request is processed in accordance with the processing request, the request accepting server executes a process according to the processing request by using a handle regarding the processing target data. When an entity of the processing target data is necessary, the entity of the processing target data is obtained on the basis of the handle and is processed.

In a parallel database system having an execution device on a front end side and a plurality of execution devices on a database operation side (hereinafter referred to as the &ldquo;divided data side&rdquo;), the front end side analyzes a definition statement of a routine such as a function or a procedure inputted from a user to create routine execution process instructions for executing the routine. Simultaneously, the front end side determines whether or not a database operation statement is contained in the routine, and creates check information indicative of the determination result. The front end side analyzes a query statement inputted from the user to create a first query execution process instruction to be executed on the front end side and a second query execution process instruction to be executed on the divided data side. When the divided data side executes the query in accordance with the second query execution process instruction, a routine containing no database operation statement in the second query statement process instruction is executed on the divided data side.

Growing community of inventors

Tama, Japan

Morihiro Iwata