Databases system for permitting concurrent indexing and reloading of data by early simulating the reload process to determine final locations of the data
First Claim
1. A computer-implemented method of reorganizing a DB2 database comprising data organized into rows and stored in a row-data set, each said row having at least one index field containing an index-key value,said computer-implemented method comprising the steps of:
- (a) for each said row of data,(1) retrieving said row of data from said row-data set;
(2) obtaining a row identifier, referred to as a RID, for said row of data by simulating a process of storing said row of data to a table space;
(3) writing said row of data to an interim row-data set;
(4) for each said index field,(A) determining the index-key value contained in the index field, and(B) writing said index-key value for said row and the RID for said row to an interim index-data set corresponding to said index field;
(b) executing a plurality of processes in parallel to;
(1) retrieve said rows of data from said interim row-data set and write said rows of data to a row-data set; and
(2) retrieve said index-key values and said RIDs from said interim index-data set, sorting index-key value/RID pairs by index-key value, and writing an index-key value/RID pair to an index-data set.
7 Assignments
0 Petitions
Accused Products
Abstract
The reorganization method of DB2 data files explores parallel processing, and asynchronous I/O to a great extent. It includes means to estimate an optimum configuration of system resources, such as storage devices (DASD devices), memory, and CPUs, etc, during reorganizations. The method mainly consists of four components, (1) concurrent indexing, (2) concurrent unloading of data file partitions, (3) efficient reloading of DB2 data pages and DB2 space maps, and (4) means to reduce access constraints to the DB2 recovery table.
163 Citations
1 Claim
-
1. A computer-implemented method of reorganizing a DB2 database comprising data organized into rows and stored in a row-data set, each said row having at least one index field containing an index-key value,
said computer-implemented method comprising the steps of: -
(a) for each said row of data, (1) retrieving said row of data from said row-data set;
(2) obtaining a row identifier, referred to as a RID, for said row of data by simulating a process of storing said row of data to a table space;(3) writing said row of data to an interim row-data set; (4) for each said index field, (A) determining the index-key value contained in the index field, and (B) writing said index-key value for said row and the RID for said row to an interim index-data set corresponding to said index field; (b) executing a plurality of processes in parallel to; (1) retrieve said rows of data from said interim row-data set and write said rows of data to a row-data set; and (2) retrieve said index-key values and said RIDs from said interim index-data set, sorting index-key value/RID pairs by index-key value, and writing an index-key value/RID pair to an index-data set.
-
Specification
-
- Resources
-
Current AssigneeBMC Software Incorporated (KKR & Co., Inc.)
-
Original AssigneeBMC Software Incorporated (KKR & Co., Inc.)
-
InventorsCunningham, William R., Hintz, Thomas E.
-
Primary Examiner(s)Lee, Thomas C.
-
Assistant Examiner(s)An, Meng-Ai T.
-
Application NumberUS07/473,663Time in Patent Office1,237 DaysField of Search395/500, 395/600, 395/800US Class Current1/1CPC Class CodesG06F 16/22 Indexing; Data structures t...Y10S 707/99942 Manipulating data structure...
