WO2002097675A1 - Method and system for improving response time of a query for a partitioned database object - Google Patents
Method and system for improving response time of a query for a partitioned database object Download PDFInfo
- Publication number
- WO2002097675A1 WO2002097675A1 PCT/US2002/016775 US0216775W WO02097675A1 WO 2002097675 A1 WO2002097675 A1 WO 2002097675A1 US 0216775 W US0216775 W US 0216775W WO 02097675 A1 WO02097675 A1 WO 02097675A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- rows
- query
- partitions
- partition
- local
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2453—Query optimisation
- G06F16/24532—Query optimisation of parallel queries
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y10—TECHNICAL SUBJECTS COVERED BY FORMER USPC
- Y10S—TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y10S707/00—Data processing: database and file management or data structures
- Y10S707/99931—Database or file accessing
- Y10S707/99932—Access augmentation or optimizing
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y10—TECHNICAL SUBJECTS COVERED BY FORMER USPC
- Y10S—TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y10S707/00—Data processing: database and file management or data structures
- Y10S707/99931—Database or file accessing
- Y10S707/99933—Query processing, i.e. searching
- Y10S707/99936—Pattern matching access
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y10—TECHNICAL SUBJECTS COVERED BY FORMER USPC
- Y10S—TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y10S707/00—Data processing: database and file management or data structures
- Y10S707/99931—Database or file accessing
- Y10S707/99937—Sorting
Definitions
- the present invention relates to the field of computer systems. More particularly, the invention relates to a method and system for improving the response time to execute a query involving a partitioned database object.
- Database management systems are computer-based systems that manage and store information.
- users store, update, and retrieve information by submitting commands to a database application responsible for maintaining the database.
- a database application responsible for maintaining the database.
- the user is located at a client station while the database application resides at a server.
- multi-tier database architectures are now common.
- the user is given access through a user interface (e.g., a web browser) at a user station, the user interface sends information requests to a middle-tier server (e.g., a web server), and the web server in turn handles the retrieval and packaging of data from one or more back-end database servers.
- the packaged information "page" is thereafter sent to the user interface for display to the user.
- Internet search engines maintain a searchable database of web sites that may be indexed in various ways. With the numbers of internet web sites explosively increasing in recent years, the amount of web site data maintained and searched by the internet search engine has correspondingly increased. The larger sets of data that must be searched to satisfy a search request could create increasingly large delays before search results are returned to the user by the search engine. Since many search engines performs a sorting operation upon the search results (e.g., based upon the number of times a search term appears on a web site), all of the responsive data in the larger body of information maintained by the search engine may be searched and sorted to ensure a correct order for the search results before any results are presented to the user. This is particularly frustrating for users that only wish to view the most relevant information from the first few pages of the search results, rather than the less useful information that often appears on the nth page returned by the search engine.
- a query may seek to access a partitioned database object.
- Partitioning in a database system generally refers to the process of decomposing an object into a greater number of relatively smaller objects. Smaller objects are often easier to manage and more efficient to search than larger objects. Thus, database systems utilize partitioning to decompose objects such as tables and indexes into smaller and more manageable pieces or "partitions.”
- the invention provide a method and mechanism to improve the response time of a query that is executed against a partitioned database object.
- One disclosed embodiment of the invention is particularly applicable to queries that involve relative ranking or ordering of results to be computed across multiple partitions of a partitioned database table. Since evaluation of predicates on a partitioned object typically involves iterating over one partition at a time, evaluating such queries normally requires significant overhead to retrieve, store, and sort data, as well as delays resulting from blocking the query results until all the partitions have been processed. An embodiment of the present invention addresses this problem by accessing only portions of the partitions during each pass through the partitions, processing only the retrieved portions of the partitions, and immediately returning results to the query. These steps are repeated until all of the results needed to satisfy the query has been provided. Further details of aspects, objects, and advantages of the invention are described below in the detailed description, drawings, and claims.
- FIG. 1 shows a process for executing a query according to an embodiment of the invention.
- Fig. 2a depicts an example database table.
- Fig. 2b shows an example partitioning scheme applied to the database table of Fig. 2a.
- Fig. 3 shows an illustrative use of the invention to execute a query according to one embodiment.
- FIGs. 4 and 5 are diagrams of system architectures with which the present invention may be implemented.
- One disclosed embodiment of the invention provides a method and mechanism to improve the response time of a query that is executed against a partitioned database table, and is particularly applicable to queries that involve relative ranking or ordering of results to be computed across multiple partitions of the partitioned database table. Since evaluation of predicates on a partitioned table typically involves iterating over one partition at a time, evaluating such queries normally requires significant overhead to retrieve, store, and sort data, as well as delays resulting from blocking the query results until all the partitions have been processed. An embodiment of the present invention addresses this problem by accessing only portions of the partitions, processing only the retrieved portions of the partitions, and immediately returning results to the query. [0016] Fig.
- FIG. 1 shows a flowchart of a process for executing a query according to an embodiment of the invention.
- the process receives a suitable query that searches for data within a partitioned database table.
- partitioning in a database system generally refers to the process of decomposing an object into a greater number of relatively smaller objects.
- each relevant partition possibly containing information responsive to the query is identified as a partition that must be searched; all other partitions not containing relevant information can be "pruned" from the search.
- the process can search all the partitions to execute the query, regardless of the relevance of any particular partition to the query terms.
- the process initially retrieves only a specified number of responsive rows from each partition, where each partition is accessed in a round-robin manner (i.e., retrieve n rows from a first partition, then another n rows from a second partition, etc.) (104).
- the query is initially executed against only a portion of each partition, and the initial set of results for each partition may contain only a small subset of the responsive rows for that partition.
- Other access procedures may also be used.
- the partitions could be accessed in a sort/merge process in a non-round-robin approach wherein a number of rows are retrieved and merged from certain partitions but are not yet consumed from other partitions.
- partitions can be accessed in parallel, thereby increasing scalability.
- the initial sets of results are returned from the partitions, where they are examined as a group (106). The entire group of results is examined to determine which rows should be immediately output to the user as the initial set of query responses.
- the process determines whether more rows must be retrieved to fully satisfy the query (108). If so, then the above steps are repeated, with another specified number of responsive rows retrieved from each partition (104), with the results from all partitions examined as a group (106), and the next set of results immediately provided to the user (110). In an embodiment, the previously retrieved rows are merged into the new group of retrieved rows to identify the specific query result rows that should be presented to the user. These process actions repeat until no more rows are to be retrieved from the partitions to satisfy the query.
- the list of identified partitions to query may be modified, even during the middle of the process, when it is recognized that one or more partitions can be pruned from the search (112 and 114). This may occur, for example, when the initial sets of rows retrieved from a partition make it clear that the partition will not contain any rows needed to fully satisfy the query.
- Salary Table 200 is a database table having a first column 202 to store userid values and a second column 204 to store salary information for each corresponding userid. Each row in salary table 200 corresponds to a distinct userid value. For many reasons, it may be desirable to decompose salary table 200 into multiple partitions. For example, if salary table 200 contains a very large number of rows, then database maintenance operations may be more efficiently performed if the salary table 200 is stored into multiple, smaller partitions.
- Fig. 2b shows an example partitioning scheme that may be imposed upon the salary table 200 of Fig. 2a.
- a "partitioning criteria" is established that separates the data in the salary table 200 based upon the first letter of the userid value for each row. All rows in salary table 200 having a userid value beginning with the letter "a" is stored in a first partition pi . Similarly, all rows in salary table 200 having a userid value beginning with the letter "b" is stored in a second partition p2, and all rows having a userid value beginning with the letter "c” is stored in a third partition p3.
- a local index is a partitioned index that is associated with data in a specific partitioned table.
- the partitioning criteria for the local index is usually the same as that for the partitioned table.
- an index 210 may exist to index the values in the salary column for partition pi.
- Many types of indexes provide a sorted order to the information referenced by the index.
- index 210 could be structured such that the index entries are sorted based the value in the salary columns of corresponding rows in partition pi .
- Any suitable index structure may be used to provide a local index 210, e.g., a B*-tree index structure.
- Local index 210 is shown in Fig.
- WHERE rownum ⁇ 5 This query requests the top four rows from the salary table 200 where the rows are sorted based upon values in the salary column of the table.
- Fig. 3 graphically illustrates how this query is processed to improve response time according to one embodiment of the invention. From left to right, the columns in Fig. 3 illustrate the progression of steps that occur to process the above query. [0028] In column 302, a specified number of rows are retrieved from each partition. The number of rows to retrieve from each partition is dependent upon the specific use to which the invention is directed. A balance can be drawn between the granularity at which results are produced, the required response time for initial results, the number of passes that may be needed to completely satisfy the query, and the overhead involved in identifying/retrieving a given row or set of rows from a partition.
- the subsequent passes can thereafter retrieve a larger numbers of rows from targeted partitions.
- a large number of rows can be retrieved in the initial polling round, with smaller numbers of rows for subsequent rounds.
- a different number of rows may be retrieved from different partitions.
- the number of rows to retrieve from each partition can equal the number of rows that must be returned to satisfy a query (i.e., retrieve n rows from each partition if the query contains the clause "WHERE rownum ⁇ n”) , which allows the process to complete after only a single round of polling.
- each partition pi, p2, and p3 corresponds to a local index 210, 212, and 214, respectively.
- Each local index corresponds to a sorted ordering, based upon salary values, for their respective partitions. Thus, the local indexes can be used to easily identify the rows in their corresponding partitions having the highest salary values.
- the two index entries 228 and 229 for the two highest salary values correspond to rows 224 and 220 in partition pi.
- the local index 212 can be used to identify rows 232 and 230 as having the highest salary values in the partition.
- local index 214 can be used to identify rows 248 and 246 as having the highest salary values in partition p3. Therefore, each identified set of rows are retrieved from partitions pi, p2, and p3 during the actions performed in column 302 of Fig. 3.
- a structure can be dynamically constructed to provide this ordering information.
- a local index can be dynamically constructed for partition p3 to provide ordering information for that partition. The cost/benefit of dynamically constructing this new local index will vary depending upon existing system conditions, such as the number of rows in each partition and the number of rows sought by the query.
- column 304 shows the rows retrieved from each partition being merged together and sorted as a group. It logically follows that if the two highest salary value rows from each partition for salary table 200 is retrieved, and the entire group of retrieved rows is sorted, then the two rows having the highest salary values for the group will also correspond to the two highest salary values for the entire salary table 200. Thus, the two rows having the highest salary values can be immediately returned to the user, as shown in column 306 of Fig. 3. Since the query calls for the four rows from salary table 200 having the highest salary values, half of the required query response is immediately being provided to the user. [0033] This highlights a significant advantage of the present invention.
- the query calls for four rows to be returned, and only two rows were provided in the actions of column 306.
- additional rows are to be retrieved from the partitions of salary table 200 to satisfy the query.
- Column 308 shows the already retrieved rows that remain after the first set of results are returned to the user.
- a determination can be made whether any partitions can be pruned from additional processing. If a given partition does not contain any rows that can possibly be retrieved and returned to satisfy the stated query conditions, then the partition can be pruned from further processing.
- the retrieved rows for partition p3 correspond to salary values of "25" and "20". Based upon the local index for partition p3, it is known that all other rows in partition p3 contain salary values that are equal to or less than these values.
- partition p3 can be pruned from additional processing.
- partition p2 cannot be pruned from additional processing, since it is possible that this partition contains rows having salary values higher than the two rows corresponding to userids Al 5 and AOl (rows 224 and 220).
- partition pi can be pruned from further processing, since local index 210 makes it clear that no additional rows in partition pi can have higher salary values than rows 224 and 220 (these are the two highest salary value rows in partition pi).
- another row in partition pi has the same salary value as row 220 (such as row 222).
- the system can be configured to either continue processing partition pi, or prune this partition pi from further processing. For the purpose of illustrating the example in Fig.
- partition pi will not be pruned from further processing.
- Two additional rows are therefore retrieved from partitions pi and p2, as shown in column 310 of Fig. 3.
- the retrieved rows correspond to the highest salary values for the remaining rows in each respective partition.
- the number of rows to retrieve in each pass remains the same as the number of rows retrieved during initial pass.
- the number of rows to retrieve may be adjusted based upon information or statistics gather during a previous pass. Some of the considerations that may be considered before adjusting the number of rows are similar to the considerations previously described with respect to selecting an initial number of rows to retrieve.
- the newly retrieved rows are merged with the previously retrieved rows that have not yet been returned as query results, and the entire set of rows is sorted as a group, as shown in column 312 of Fig. 3.
- the two rows having the highest salary values correspond to the highest salary values for all remaining rows in salary table 200.
- the two rows having the highest salary values (having userids of A15 and B60) are provided as the next set of results to the query, as shown in column 314. Since a total of four rows has been provided, the query is now fully satisfied.
- the query includes a WHERE clause. It is noted that the present invention is usable to improve response time and provides performance benefits even if the type of WHERE clause shown in the example query is not present.
- the present invention can also be applied to database systems that employ user- defined indexes and ancillary operators.
- Ancillary operators involve a class of database operators for which data (“ancillary data") may be shared between operations.
- ancillary data data
- a context object is defined to store data from a first operator, which is thereafter usable by a related ancillary operator to share data within the context object.
- the contains( ) function is an operator that accepts two parameters Ol and O2 (Ol corresponds to "resume” and O2 corresponds to "Biology" in this example query).
- the contains( ) function returns a True/False flag that indicates whether the entity represented by the Ol parameter contains the text of the value in the 02 parameter.
- the rank( ) function is an ancillary operator that ranks the various rows in the relative order of significance to the contains( ) operator. Since rank( ) and contains( ) can be configured as related operators, common ancillary data can be accessed between these operators.
- this query seeks the top 100 ranked rows in the Student table that is responsive to the contains() operator/predicate in the WHERE clause. Assume that the Student table is decomposed into ten partitions, and a suitable local user-defined index exists for each partition.
- One approach to evaluating this type of query in a database system is to evaluate the contains( ) operator for each partition, store the results until all partitions have been evaluated, rank the collected results for all the partitions, and then return the top 100 rows from the ranked results.
- the drawbacks with such an approach have been described in detail above, including excessive overhead consumption for retrieving, storing, and sorting rows from all partitions, and blocking all results until all partitions have been processed, even though only 100 rows need to be returned.
- the server pushes down the evaluation of the rank( ) operator to each individual partition in addition to the contains( ) operator, and the results are polled from each partition.
- Each partition returns a specified number of rows that have been ranked within that partition.
- the server polls all the partitions and collects the respective results from the partitions. After the first round of the polling, the server can return a subset of the result rows to the user and decide if additional polling should be performed. If it does, polling is performed to return another result set of rows from the partitions. As described above, a subset of partitions can be eliminated from the polling after every round of polling is complete. [0044] If the cost of evaluating rank( ) is relatively high, then fewer rows are retrieved during each polling round, according to one embodiment.
- the inventive process further comprises a step to identify specific queries that may benefit from the invention, and only applying the invention to these identified queries.
- the following are examples of characteristics that may be used, whether in combination or separately, to identify such queries: (a) queries that should be optimized for response time rather than total throughput; (b) queries having a sorting or ordering element (e.g., having an "ORDER BY" clause); (c) queries against a partitioned objects; (d) an index or other structure is already available to provide ordering information for partitions being queried; and, (e) queries that seek to limit the number of responses (e.g., using a "WHERE rownum ⁇ «" clause).
- a computer system 420 includes a host computer 422 connected to a plurality of individual user stations 424.
- the user stations 424 each comprise suitable data terminals, for example, but not limited to, e.g., personal computers, portable laptop computers, or personal data assistants ("PDAs"), which can store and independently run one or more applications, i.e., programs.
- PDAs personal data assistants
- some of the user stations 424 are connected to the host computer 422 via a local area network (“LAN”) 426.
- LAN local area network
- Other user stations 424 are remotely connected to the host computer 422 via a public telephone switched network (“PSTN”) 428 and/or a wireless network 430.
- PSTN public telephone switched network
- the host computer 422 operates in conjunction with a data storage system 431, wherein the data storage system 431 contains a database 432 that is readily accessible by the host computer 422.
- a multiple tier architecture can be employed to connect user stations 424 to a database 432, utilizing for example, a middle application tier (not shown).
- the database 432 may be resident on the host computer, stored, e.g., in the host computer's ROM, PROM, EPROM, or any other memory chip, and/or its hard disk.
- the database 432 may be read by the host computer 422 from one or more floppy disks, flexible disks, magnetic tapes, any other magnetic medium, CD-ROMs, any other optical medium, punchcards, papertape, or any other physical medium with patterns of holes, or any other medium from which a computer can read.
- the host computer 422 can access two or more databases 432, stored in a variety of mediums, as previously discussed. [0048] Referring to Fig. 5, in an embodiment, each user station 424 and the host computer 422, each referred to generally as a processing unit, embodies a general architecture 505.
- a processing unit includes a bus 506 or other communication mechanism for communicating instructions, messages and data, collectively, information, and one or more processors 507 coupled with the bus 506 for processing information.
- a processing unit also includes a main memory 508, such as a random access memory (RAM) or other dynamic storage device, coupled to the bus 506 for storing dynamic data and instructions to be executed by the processor(s) 507.
- the main memory 508 also may be used for storing temporary data, i.e., variables, or other intermediate information during execution of instructions by the processor(s) 507.
- a processing unit may further include a read only memory (ROM) 509 or other static storage device coupled to the bus 506 for storing static data and instructions for the processor(s) 507.
- ROM read only memory
- a storage device 510 such as a magnetic disk or optical disk, may also be provided and coupled to the bus 506 for storing data and instructions for the processor(s) 507.
- a processing unit may be coupled via the bus 506 to a display device 511, such as, but not limited to, a cathode ray tube (CRT), for displaying information to a user.
- a display device 511 such as, but not limited to, a cathode ray tube (CRT)
- An input device 512 is coupled to the bus 506 for communicating information and command selections to the processor(s) 507.
- Another type of user input device may include a cursor control 513, such as, but not limited to, a mouse, a trackball, a fingerpad, or cursor direction columns, for communicating direction information and command selections to the processor(s) 507 and for controlling cursor movement on the display 511.
- the individual processing units perform specific operations by their respective processor(s) 507 executing one or more sequences of one or more instructions contained in the main memory 508.
- Such instructions may be read into the main memory 508 from another computer-usable medium, such as the ROM 509 or the storage device 510.
- Execution of the sequences of instructions contained in the main memory 508 causes the processor(s) 507 to perform the processes described herein.
- hard-wired circuitry may be used in place of or in combination with software instructions to implement the invention.
- embodiments of the invention are not limited to any specific combination of hardware circuitry and/or software.
- Non-volatile media i.e., media that can retain information in the absence of power
- Volatile media i.e., media that can not retain information in the absence of power
- Transmission media includes coaxial cables, copper wire and fiber optics, including the wires that comprise the bus 506.
- Transmission media can also take the form of carrier waves; i.e., electromagnetic waves that can be modulated, as in frequency, amplitude or phase, to transmit information signals.
- transmission media can take the form of acoustic or light waves, such as those generated during radio wave and infrared data communications.
- Common forms of computer-usable media include, for example: a floppy disk, flexible disk, hard disk, magnetic tape, any other magnetic medium, CD-ROM, any other optical medium, punchcards, papertape, any other physical medium with patterns of holes, RAM, ROM, PROM (i.e., programmable read only memory), EPROM (i.e., erasable programmable read only memory), including FLASH-EPROM, any other memory chip or cartridge, carrier waves, or any other medium from which a processor 507 can retrieve information.
- Each processing unit may also include a communication interface 514 coupled to the bus 506.
- the communication interface 514 provides two-way communication between the respective user stations 524 and the host computer 522.
- the communication interface 514 of a respective processing unit transmits and receives electrical, electromagnetic or optical signals that include data streams representing various types of information, including instructions, messages and data.
- a communication link 515 links a respective user station 524 and a host computer 522.
- the communication link 515 may be a LAN 426, in which case the communication interface 514 may be a LAN card.
- the communication link 515 may be a PSTN 428, in which case the communication interface 514 may be an integrated services digital network (ISDN) card or a modem.
- ISDN integrated services digital network
- the communication link 515 may be a wireless network 430.
- a processing unit may transmit and receive messages, data, and instructions, including program, i.e., application, code, through its respective communication link 515 and communication interface 514. Received program code may be executed by the respective processor(s) 507 as it is received, and/or stored in the storage device 510, or other associated non- volatile media, for later execution. In this manner, a processing unit may receive messages, data and/or program code in the form of a carrier wave.
Abstract
Description
Claims
Priority Applications (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2003500785A JP4246623B2 (en) | 2001-05-31 | 2002-05-28 | Method and system for improving query response time for partitioned database objects |
AU2002312104A AU2002312104B2 (en) | 2001-05-31 | 2002-05-28 | Method and system for improving response time of a query for a partitioned database object |
EP02739458A EP1395927A4 (en) | 2001-05-31 | 2002-05-28 | Method and system for improving response time of a query for a partitioned database object |
CA002448730A CA2448730A1 (en) | 2001-05-31 | 2002-05-28 | Method and system for improving response time of a query for a partitioned database object |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US09/872,670 | 2001-05-31 | ||
US09/872,670 US6795817B2 (en) | 2001-05-31 | 2001-05-31 | Method and system for improving response time of a query for a partitioned database object |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2002097675A1 true WO2002097675A1 (en) | 2002-12-05 |
Family
ID=25360075
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/US2002/016775 WO2002097675A1 (en) | 2001-05-31 | 2002-05-28 | Method and system for improving response time of a query for a partitioned database object |
Country Status (6)
Country | Link |
---|---|
US (1) | US6795817B2 (en) |
EP (1) | EP1395927A4 (en) |
JP (1) | JP4246623B2 (en) |
AU (1) | AU2002312104B2 (en) |
CA (1) | CA2448730A1 (en) |
WO (1) | WO2002097675A1 (en) |
Families Citing this family (38)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6434567B1 (en) * | 1996-07-30 | 2002-08-13 | Carlos De La Huerga | Method for specifying enterprise-wide database address formats |
CA2363187A1 (en) * | 2001-11-19 | 2003-05-19 | Ibm Canada Limited-Ibm Canada Limitee | Index sampled tablescan |
US6907422B1 (en) * | 2001-12-18 | 2005-06-14 | Siebel Systems, Inc. | Method and system for access and display of data from large data sets |
US7007024B2 (en) * | 2002-03-29 | 2006-02-28 | Panasas, Inc. | Hashing objects into multiple directories for better concurrency and manageability |
US7356523B2 (en) * | 2002-05-23 | 2008-04-08 | International Business Machines Corporation | Dynamic optimization of prepared statements in a statement pool |
US7127463B2 (en) * | 2002-04-18 | 2006-10-24 | International Business Machines Corporation | Optimization of database network traffic based upon data-use analysis |
WO2004021152A2 (en) * | 2002-08-30 | 2004-03-11 | Findwhat.Com | System and method for pay for performance advertising employing multiple sets of advertisement listings |
US20040111388A1 (en) * | 2002-12-06 | 2004-06-10 | Frederic Boiscuvier | Evaluating relevance of results in a semi-structured data-base system |
US20040236724A1 (en) * | 2003-05-19 | 2004-11-25 | Shu-Yao Chien | Searching element-based document descriptions in a database |
US7464070B2 (en) * | 2003-07-29 | 2008-12-09 | Hitachi, Ltd. | Database query operations using storage networks |
US20050171875A1 (en) * | 2004-01-30 | 2005-08-04 | Michael Seitz | System and method for synchronizing commitment updates |
US7447679B2 (en) * | 2004-05-07 | 2008-11-04 | Oracle International Corporation | Optimizing execution of a database query by using the partitioning schema of a partitioned object to select a subset of partitions from another partitioned object |
US7962489B1 (en) * | 2004-07-08 | 2011-06-14 | Sage-N Research, Inc. | Indexing using contiguous, non-overlapping ranges |
US20060167845A1 (en) * | 2005-01-25 | 2006-07-27 | International Business Machines Corporation | Selection of optimal plans for FIRST-N-ROW queries |
US7725468B2 (en) * | 2005-04-08 | 2010-05-25 | Oracle International Corporation | Improving efficiency in processing queries directed to static data sets |
US20060242119A1 (en) * | 2005-04-20 | 2006-10-26 | Bea Systems, Inc. | SQL-style filtered rowset |
JP4839706B2 (en) * | 2005-07-12 | 2011-12-21 | 株式会社日立製作所 | Index management method for database management system |
US20070113031A1 (en) * | 2005-11-16 | 2007-05-17 | International Business Machines Corporation | Memory management system and method for storing and retrieving messages |
US20080222107A1 (en) * | 2006-07-21 | 2008-09-11 | Maluf David A | Method for Multiplexing Search Result Transmission in a Multi-Tier Architecture |
US7831620B2 (en) * | 2006-08-31 | 2010-11-09 | International Business Machines Corporation | Managing execution of a query against a partitioned database |
US20080071744A1 (en) * | 2006-09-18 | 2008-03-20 | Elad Yom-Tov | Method and System for Interactively Navigating Search Results |
US7953083B1 (en) | 2006-12-12 | 2011-05-31 | Qurio Holdings, Inc. | Multicast query propagation scheme for a peer-to-peer (P2P) network |
US7743013B2 (en) * | 2007-06-11 | 2010-06-22 | Microsoft Corporation | Data partitioning via bucketing bloom filters |
US8201171B2 (en) * | 2007-06-27 | 2012-06-12 | Microsoft Corporation | Adjacent data parallel and streaming operator fusion |
US8886803B1 (en) * | 2007-09-20 | 2014-11-11 | Qurio Holdings, Inc. | Dynamic query response in a distributed network |
US7987164B2 (en) * | 2008-01-09 | 2011-07-26 | Oracle International Corporation | Method and system for speeding up rebuild of user-defined indexes during partition maintenance operations in the database systems |
US20100191717A1 (en) * | 2009-01-28 | 2010-07-29 | Goetz Graefe | Optimization of query processing with top operations |
EP2290562A1 (en) * | 2009-08-24 | 2011-03-02 | Amadeus S.A.S. | Segmented main-memory stored relational database table system with improved collaborative scan algorithm |
US20110153650A1 (en) * | 2009-12-18 | 2011-06-23 | Electronics And Telecommunications Research Institute | Column-based data managing method and apparatus, and column-based data searching method |
US8959094B2 (en) * | 2010-05-28 | 2015-02-17 | Oracle International Corporation | Early return of partial sort results in a database system |
US9449045B2 (en) * | 2011-04-08 | 2016-09-20 | Sybase, Inc. | System and method for enhanced query optimizer search space ordering |
US8918436B2 (en) * | 2011-12-22 | 2014-12-23 | Sap Ag | Hybrid database table stored as both row and column store |
IN2013MU02853A (en) | 2013-09-02 | 2015-07-03 | Tata Consultancy Services Ltd | |
US10108649B2 (en) * | 2014-02-25 | 2018-10-23 | Internatonal Business Machines Corporation | Early exit from table scans of loosely ordered and/or grouped relations using nearly ordered maps |
KR101679838B1 (en) | 2015-04-27 | 2016-11-25 | 주식회사 알티베이스 | Array sequencer for dynamic sequencing in database |
US10331645B2 (en) * | 2015-08-17 | 2019-06-25 | International Business Machines Corporation | Data-partitioning for processing loosely ordered relations |
US11138230B2 (en) * | 2018-03-26 | 2021-10-05 | Mcafee, Llc | Methods, apparatus, and systems to aggregate partitioned computer database data |
US11468102B2 (en) * | 2020-06-05 | 2022-10-11 | Teradata Us, Inc. | Optimizing limit queries over analytical functions |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6128610A (en) * | 1996-07-09 | 2000-10-03 | Oracle Corporation | Index with entries that store the key of a row and all non-key values of the row |
US6192358B1 (en) * | 1997-07-10 | 2001-02-20 | Internatioanal Business Machines Corporation | Multiple-stage evaluation of user-defined predicates |
US6219662B1 (en) * | 1997-07-10 | 2001-04-17 | International Business Machines Corporation | Supporting database indexes based on a generalized B-tree index |
US6338056B1 (en) * | 1998-12-14 | 2002-01-08 | International Business Machines Corporation | Relational database extender that supports user-defined index types and user-defined search |
Family Cites Families (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5813000A (en) * | 1994-02-15 | 1998-09-22 | Sun Micro Systems | B tree structure and method |
US5960194A (en) * | 1995-09-11 | 1999-09-28 | International Business Machines Corporation | Method for generating a multi-tiered index for partitioned data |
USRE37965E1 (en) * | 1995-09-27 | 2003-01-07 | International Business Machines Corporation | Method for localizing execution or subqueries and determining collocation of execution of subqueries in a parallel database |
US5710915A (en) * | 1995-12-21 | 1998-01-20 | Electronic Data Systems Corporation | Method for accelerating access to a database clustered partitioning |
US6510435B2 (en) * | 1996-09-02 | 2003-01-21 | Rudolf Bayer | Database system and method of organizing an n-dimensional data set |
JP3836928B2 (en) * | 1997-02-26 | 2006-10-25 | 株式会社日立製作所 | Database processing method |
US6405198B1 (en) * | 1998-09-04 | 2002-06-11 | International Business Machines Corporation | Complex data query support in a partitioned database system |
US6401083B1 (en) * | 1999-03-18 | 2002-06-04 | Oracle Corporation | Method and mechanism for associating properties with objects and instances |
US6370522B1 (en) * | 1999-03-18 | 2002-04-09 | Oracle Corporation | Method and mechanism for extending native optimization in a database system |
US6665684B2 (en) * | 1999-09-27 | 2003-12-16 | Oracle International Corporation | Partition pruning with composite partitioning |
US6546382B1 (en) * | 1999-11-03 | 2003-04-08 | Oracle Corporation | Finding the TOP N values through the execution of a query |
US6421666B1 (en) * | 1999-11-08 | 2002-07-16 | Oracle Corporation | Mechanism for sharing ancillary data between a family of related functions |
US6470331B1 (en) * | 1999-12-04 | 2002-10-22 | Ncr Corporation | Very large table reduction in parallel processing database systems |
US6564221B1 (en) * | 1999-12-08 | 2003-05-13 | Ncr Corporation | Random sampling of rows in a parallel processing database system |
-
2001
- 2001-05-31 US US09/872,670 patent/US6795817B2/en not_active Expired - Lifetime
-
2002
- 2002-05-28 JP JP2003500785A patent/JP4246623B2/en not_active Expired - Lifetime
- 2002-05-28 EP EP02739458A patent/EP1395927A4/en not_active Ceased
- 2002-05-28 CA CA002448730A patent/CA2448730A1/en not_active Abandoned
- 2002-05-28 AU AU2002312104A patent/AU2002312104B2/en not_active Expired
- 2002-05-28 WO PCT/US2002/016775 patent/WO2002097675A1/en active Application Filing
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6128610A (en) * | 1996-07-09 | 2000-10-03 | Oracle Corporation | Index with entries that store the key of a row and all non-key values of the row |
US6192358B1 (en) * | 1997-07-10 | 2001-02-20 | Internatioanal Business Machines Corporation | Multiple-stage evaluation of user-defined predicates |
US6219662B1 (en) * | 1997-07-10 | 2001-04-17 | International Business Machines Corporation | Supporting database indexes based on a generalized B-tree index |
US6338056B1 (en) * | 1998-12-14 | 2002-01-08 | International Business Machines Corporation | Relational database extender that supports user-defined index types and user-defined search |
Non-Patent Citations (4)
Title |
---|
CAREY M J, ON SAYING 'ENOUGH ALREADY!' IN SQL |
CAREY M J, REDUCING THE BREAKING DISTANCE OF AN SQL QUERY ENGINE |
CHAUDHURI S, EVALUATING TOP-K SELECTION QUERIES |
KIAN-LEE TAN, ON GETTING SOME ANSWERS QUICKLY, AND PERHAPS MORE LATER |
Also Published As
Publication number | Publication date |
---|---|
JP2004528660A (en) | 2004-09-16 |
CA2448730A1 (en) | 2002-12-05 |
AU2002312104B2 (en) | 2008-04-10 |
EP1395927A4 (en) | 2007-04-25 |
US6795817B2 (en) | 2004-09-21 |
US20020184253A1 (en) | 2002-12-05 |
JP4246623B2 (en) | 2009-04-02 |
EP1395927A1 (en) | 2004-03-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
AU2002312104B2 (en) | Method and system for improving response time of a query for a partitioned database object | |
AU2002312104A1 (en) | Method and system for improving response time of a query for a partitioned database object | |
Chu et al. | A transaction-based approach to vertical partitioning for relational database systems | |
US7984039B2 (en) | Merging of results in distributed information retrieval | |
US7158996B2 (en) | Method, system, and program for managing database operations with respect to a database table | |
US6581055B1 (en) | Query optimization with switch predicates | |
DeWitt et al. | Parallel sorting on a shared-nothing architecture using probabilistic splitting | |
US5913207A (en) | Database system index selection using index configuration enumeration for a workload | |
US9298761B2 (en) | Adaptive merging in database indexes | |
US7567952B2 (en) | Optimizing a computer database query that fetches n rows | |
US6266660B1 (en) | Secondary index search | |
US8935231B2 (en) | Optimizing a query to a partitioned database table using a virtual maintained temporary index that spans multiple database partitions | |
Cheung et al. | Constructing suffix tree for gigabyte sequences with megabyte memory | |
US5913206A (en) | Database system multi-column index selection for a workload | |
IL147694A (en) | Database system for viewing effects of changes to an index for a query optimization | |
Leslie et al. | Efficient search of multi-dimensional b-trees. | |
Bachhav et al. | An efficient query optimizer with materialized intermediate views in distributed and cloud environment | |
MahmoudiNasab et al. | AdaptRDF: adaptive storage management for RDF databases | |
US20100268723A1 (en) | Method of partitioning a search query to gather results beyond a search limit | |
US6694324B1 (en) | Determination of records with a specified number of largest or smallest values in a parallel database system | |
Bovey et al. | An algorithm for weighted searching on a Boolean system | |
Gibas et al. | Online index recommendations for high-dimensional databases using query workloads | |
JPH10269225A (en) | Data base dividing method | |
Richards | The Advancing Art of Indexing | |
Damasceno et al. | Tinycubes: A modular technology for interactive visual analysis of historical and continuously updated spatiotemporal data |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AK | Designated states |
Kind code of ref document: A1 Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BY BZ CA CH CN CO CR CU CZ DE DK DM DZ EC EE ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NO NZ OM PH PL PT RO RU SD SE SG SI SK SL TJ TM TN TR TT TZ UA UG UZ VN YU ZA ZM ZW |
|
AL | Designated countries for regional patents |
Kind code of ref document: A1 Designated state(s): GH GM KE LS MW MZ SD SL SZ TZ UG ZM ZW AM AZ BY KG KZ MD RU TJ TM AT BE CH CY DE DK ES FI FR GB GR IE IT LU MC NL PT SE TR BF BJ CF CG CI CM GA GN GQ GW ML MR NE SN TD TG |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
WWE | Wipo information: entry into national phase |
Ref document number: 2002312104 Country of ref document: AU Ref document number: 1532/KOLNP/2003 Country of ref document: IN |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2448730 Country of ref document: CA |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2003500785 Country of ref document: JP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2002739458 Country of ref document: EP |
|
WWP | Wipo information: published in national office |
Ref document number: 2002739458 Country of ref document: EP |
|
REG | Reference to national code |
Ref country code: DE Ref legal event code: 8642 |