Online reorganization and redefinition of relational database tables

[0001] The present invention relates generally to relational databases and, more specifically, to online reorganization and redefinition of relational database tables.

[0002] Data containers in large active relational databases often need to be reorganized and/or redefined. For example, a particular relational database table may be redefined to improve the performance of processing queries and data manipulation language (DML) operations against the particular relational table. The redefinition operation may involve, for example, adding a new column to the table. Tables may also have to be reorganized because the database and business applications that use the tables may require changes to the underlying structure of the tables. Tables that are to be reorganized and/or redefined are herein referred to as “target tables”.

[0003] In one approach, a target table is reorganized and/or redefined by locking the target table in a manner that prevents users from querying and performing DML operations on all or part of the target table. In other words, the target table is inaccessible to users for the entire duration of the reorganization and/or redefinition operation. The duration of inaccessibility to the data in the relational database is herein referred to as “downtime”. The problem with such an approach is that the downtime associated with the reorganization/redefinition may be unacceptably lengthy if the target tables are massive and the reorganization/redefinition is extensive. In addition, it is difficult to implement such a feature when dependent objects, such as indexes, are also involved.

[0004] In the conventional approach, the target table is taken “offline” for the purpose of reorganization/redefinition. Thus, the target table is inaccessible to users during the period that the target is offline. For example, a given target table is first exported from the relational database for the process of reorganization/redefinition. After the target table is reorganized/redefined, the target table is imported back into the relational database. One problem with such an approach is that indexes, constraints, and triggers that are associated with a target table are typically dropped in order to speed up the import process. Thus, the indexes, constraints, and triggers associated with the target table need to be re-created. Each hour of downtime can cost millions of dollars for a corporation.

[0005] Based on the foregoing, it is clearly desirable to provide a mechanism for reorganizing/redefining data in relational databases with little or no downtime.

[0006] Techniques are provided for allowing the reorganization and/or redefinition of data in a relational database with minimal or no downtime by allowing “online” reorganization/redefinition of the data. According to one aspect of the invention, an empty “revised” table is created in the same schema as the target table. Initially, the revised table is merely an “interim” table that is not directly accessed by user queries. However, when completed and populated, the revised table is used to answer queries that are directed to the target table.

[0007] Specifically, the revised table is created with the attributes that are desired for the reorganization/redefinition of the target table. Triggers, constraints, grants and indexes that are associated with the target table are recreated for the revised table. The revised table is populated with data from the target table using an incremental refresh mechanism, similar to the mechanisms commonly used to incrementally refresh materialized views.

[0008] According to one feature, the revised table is initialized and periodically synchronized with the target table. After the reorganization and/or redefinition is complete on the revised table, the database system begins using the revised table to answer queries that are directed to the target table. According to one embodiment, the revised table and the target table switch identities such that the revised table is assumes the role of the target table.

[0012] Techniques are provided for online reorganization and redefinition of relational database tables. In the following description, for the purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the present invention. It will be apparent, however, to one skilled in the art that the present invention may be practiced without these specific details. In other instances, well-known structures and devices are shown in block diagram form in order to avoid unnecessarily obscuring the present invention.

[0013] The feature for allowing users full access to a target table in a relational database during the process of data reorganization and/or redefinition is herein referred to as “online modification” of the target table. During the online modification of the target table, users can access the target table, both for read (select) and write (insert and modify) operations. Typically, during an upgrade of a given application of the relational database system, the database administrator needs to modify one or more tables in the relational database. By using the online modification technique, the database administrator can perform the online modification operation at any time without impinging on the data availability of the target tables.

[0014] The online modification technique allows the database administrator to perform various types of modifications to the target table, including but not limited to: 1) modify the storage parameters of the target table, 2) add support for parallel queries to be performed on the target table after the target table is modified, 3) change a heap organized target table to an Index Only Table (IOT) or vice versa, 4) move the target table to a different tablespace, 5) add or drop partitioning support for the target table, 6) recreate the target table in order to reduce fragmentation, 7) add, drop, or rename columns in the target table, 8) transform data in the target table.

[0015] FIG. 1 is a flowchart that illustrates the online modification operation of a target table in a relational database. At block 102 of FIG. 1, an empty revised table is created with all the attributes that are desired for the online modification. At block 104, the database system starts populating the revised table based on data in the target table. Data from the target table is copied over to the revised table based on column mapping information. Column mapping information is explained in greater detail herein.

[0016] At block 106, any triggers, indexes, grants and constraints that exist for the target table are created for the revised table. However, such triggers, indexes, grants and constraints are created disabled for the revised table. According to one embodiment, such triggers, indexes, grants and constraints are enabled only when the online modification operation is successfully completed.

[0017] At block 108, the data in the revised table is synchronized with the data in the target table. During part of this particular step, access to the target table may be disabled. During all other steps of the online modification process, the target is accessible by users. After step 108, the revised table will accurately reflect all changes made to data in the target table, including those changes that were made during the online modification process.

[0018] At block 110, all subsequent attempts to access the target table are caused to access the revised table instead. Example techniques for accomplishing step 110 are described hereafter. However, the present invention is not limited to any specific technique for causing subsequent attempt to access the target table to access the revised table.

[0019] For the purpose of explanation, many of the examples given herein address the scenario in which a single target table is being reorganized into a single revised table. However, the present invention is not limited to this scenario. Specifically, the online modification techniques described herein can be used for transforming: 1) one table to many tables, 2) many tables to one table, 3) many tables to many tables, and 4) sub-table to sub-table. Further details are described herein.

[0020] Start of Online Modification Operation

[0021] According to certain embodiments, the online modification operation is started by providing the following as inputs: 1) the schema name of the target table, 2) the name of the target table, 3) the name of the revised table, and 4) column mapping information between the target table and the revised table. In certain embodiments, if the column mapping information is not supplied, then it is assumed that all the columns of the target table, with the names of the columns unchanged, are to be included in the revised table. If the column mapping information is supplied, then only those columns that are explicitly specified in the column mapping information are considered.

[0022] Intermediate Synchronization

[0023] Because the target table remains online, it is possible that a large number of Data Manipulation Language (DML) operations have occurred on the target table between the time when the online modification operation is started and the time when the online modification is completed. Only the final step of the online modification operation involves locking the target table to synchronize the data in the revised table with the data in the target table.

[0024] In certain embodiments of the invention, the synchronization of data between the target table and the revised table takes place periodically during the entire online modification operation. The synchronization of data between the target table and the revised table may be performed multiple times according to the discretion of the database administrator who is tasked with performing the online modification. By periodically synchronizing the data between the target table and the revised table, the time taken to complete a final synchronization immediately prior to bringing the revised table online is greatly reduced.

[0025] The data in the revised table can be validated to ensure that the online modifications are performed accurately.

[0026] Various techniques may be used to perform the synchronization of data between the target table and the revised table. Such techniques may use similar or identical techniques to those currently used to perform incremental refreshes to materialized views in relational database systems. Examples of such techniques are described in U.S. Pat. No. 6,205,451, entitled, “METHOD AND APPARATUS FOR INCREMENTAL REFRESH OF SUMMARY TABLES IN A DATABASE SYSTEM” filed by William D. Norcott and James Finnerty on May 22, 1998, the contents of which are incorporated by reference herein, in U.S. patent application Ser. No. 09/224,063, entitled, “METHOD AND APPARATUS FOR EFFICIENTLY REFRESHING SETS OF SUMMARY TABLES AND MATERIALIZED VIEWS IN A DATABASE MANAGEMENT SYSTEM” filed by William D. Norcott and Mohamed Ziauddin on Dec. 28, 1998, and in U.S. Pat. No. 5,963,959, entitled “FAST REFRESH OF SNAPSHOTS” filed by “Harry Sun, Alan Downing, and Benny Souder” the contents of which are incorporated by reference herein.

[0027] Completion of Online Modification Operation

[0028] Invoking the appropriate database procedure completes the online modification operation. According to certain embodiments, when the creation of the revised table is complete, i.e., when the revised table contains the online modifications that are desired for the target table, a mechanism is provided to cause all subsequent user attempts to access the target table to access to revised table instead.

[0029] The operation of causing all subsequent user attempts to access the target table to access to revised table instead may be implemented in a variety of ways, including what is herein referred to as the “switch operation”. The switch operation is very brief and has a duration that is independent of the size of the table and independent of the complexity of the redefinition. According to certain embodiments of the invention, the switch operation involves switching the identity of the target table with the identity of the revised table.

[0030] Due to the switch operation, the revised table assumes the identity of the target table and has all the attributes and modifications that are desired. The revised table that assumes the identity of the target table due to the switch operation will herein be referred to as the “online revised table”. The indexes, constraints, grants and triggers are enabled for the online revised table. The indexes in the online revised table may now be re-named if so desired.

[0031] Similarly, due to the switch operation, the target table assumes the identity of the revised table and is referred to as the offline target table. The offline target table can be dropped from the database. Any indexes, constraints, grants and triggers associated with the new offline target table are disabled and/or dropped.

[0032] Abort and Clean Up

[0033] In the event that an error occurs during the online modification operation, or if the database administrator chooses to abort the online table modification operation, the appropriate database procedure for aborting the online table modification operation can be invoked. When invoked, such a procedure drops temporary logs and tables that are associated with the online modification operation. Once such a procedure is invoked, the database administrator can then drop the revised table and the objects that are associated with the revised table.

[0034] Example of Online Table Modification (One Table to One Table)

[0035] The following example illustrates online modification of a non-partitioned table named EMP, and with columns EMPNO, NAME, SALARY, PHONE. Assume that target table EMP is in schema U1. Target table EMP is to be online modified as follows: 1) Column salary is to be multiplied by a factor of 1.10 and renamed as SAL, 2) Column PHONE is to be dropped, 3) a new column DEPTNO with default value of 10 is to be added, 4) the modified table is to be partitioned by range on EMPNO. The following steps 1 through 4 illustrate the online modification of table Emp. The syntax used in the statements in Steps 1 through 4 is merely illustrative. The actual syntax of statements used for online table modification may vary from implementation to implementation. The present invention is not limited to any particular syntax.

[0036] STEP 1: Create a Revised Table int_—emp

• [0037] CREATE TABLE int_—emp
• [0038] (empno NUMBER PRIMARY KEY, name VARCHAR2(100), sal NUMBER, deptno NUMBER DEFAULT 10)
• [0039] PARTITION BY RANGE(empno)
• [0040] PARTITION emp1000 VALUES LESS THAN (1000) TABLESPACE tbs_—1,
• [0041] PARTITION emp2000 VALUES LESS THAN (2000) TABLESPACE tbs_—2;

[0042] STEP 2: Start the Online Modification Operation

• [0043] DBMS_—REDEFINITION.START_—REDEF_—TABLE(‘u1’, ‘emp’, ‘int_—emp’, ‘empno empno, name name, salary*1.10 sal’);

[0044] During step 2, the following occur:

• • [0045] 1) A materialized view log on EMP is created;
  • [0046] 2) A materialized view INT_—EMP is created.

[0047] In addition, any triggers, indexes, grants and constraints that exist for the target table EMP are created for the revised table INT_—EMP. However, such triggers, indexes, grants and constraints are created disabled for the revised table INT_—EMP.

[0048] STEP 3: Optionally, Synchronize the Revised Table int_—emp with Target Table Emp

• [0049] DBMS_—REDEFINITION.SYNC_—INTERIM_—TABLE(‘u1’, ‘emp’, ‘int_—emp’);

[0050] STEP 4: Complete the Online Modification

• [0051] DBMS_—REORG.FINISH_—REDEF_—TABLE(‘u1’, ‘emp’, ‘int_—emp’);

[0052] During step 4, the following occur:

• • [0053] 1) The materialized view INT_—EMP is refreshed;
  • [0054] 2) The target table EMP is locked;
  • [0055] 3) After the target table EMP has been locked, the materialized view INT_—EMP is refreshed once again;
  • [0056] 4) The identities of this table EMP and INT_—EMP are switched and any indexes, constraints, grants and triggers on the online revised table are enabled while any indexes, constraints, grants and triggers associated with the offline target table are disabled and/or dropped;
  • [0057] 5) The materialized view INT_—EMP and the materialized view logs on EMP are dropped;
  • [0058] 6) The table EMP is unlocked.

[0059] After step 4, online revised table emp is modified such that it has all the attributes of the pre-switch revised table int_—emp. The state of the pre-switch revised table is the state immediately prior to the switch operation explained herein.

[0060] Assume that in the year 2001, a company has an application that heavily uses a table named “table A” and that has the following fields:

• [0061] 1. message #
• [0062] 2. originating id
• [0063] 3. message type
• [0064] 4. text(4000)

[0065] Assume table A has a message type field. This message type field tells the application how to interpret the text information. For example, the text information could be an address or a comment. Table A is said to be in 2nd normal form. The 2nd normal form may be used by an application as a tradeoff between speed, storage space, and complexity of management and application development.

[0066] Assume that by the year 2002, the application gets more complex and the company gets larger, the application may find that the 2nd normal form provides the wrong tradeoffs. Thus, in the 2002 version of the application, target table A is modified into in the 3rd normal form. In the 3^rd normal form, target table A is split into many tables. Assume that target table A is modified into a new table A (msg#, originating id, msg_—type), table A2 (msg#, address), and table A3 (msg#, comment). The application upgrade from the 2001 to 2002 versions needs to split target table A into a new table A, and tables A2, and A3. Assume that the company does not want to suffer the costs of a lengthy downtime. Thus, table A is a good candidate for then online table modification operation.

[0067] The first step for modifying table A into a new table A, and tables A2, and A3 is to create 3 new tables as follows:

• • [0068] create table Temp_—A (msg# number, orig_—id number, msg_—type number);
  • [0069] create table A2(msg# number, address varchar2(4000));
  • [0070] create table A3msg# number, comment varchar2(4000));

[0071] Next, a query list is created:

• • [0072] Table Temp_—A is defined as: select msg#, orig_—id, msg_—type from A;
  • [0073] Table A2 is defined as: select msg#, text from A where msg_—type=1;
  • [0074] Table A3 is defined as: select msg#, text from A where msg_—type=2;
  • [0075] The list of post-modified tables are Table A, A2, A3. The list of “revised tables” are Temp_—A, A2, A3. Assume that all tables are owned by user U.

[0076] The online modification may be started by performing the following initializations:

• • [0077] owner_—list(1):=‘U’;
  • [0078] owner_—list(2):=‘U’;
  • [0079] owner_—list(3):=‘U’;
  • [0080] interim_—table_—list(1):=‘Temp_—A’;
  • [0081] interim_—table_—list(2):=‘A2 ’;
  • [0082] interim_—table_—list(3):=‘A3’;
  • [0083] post_—modified_—table_—list(1)=‘A’;
  • [0084] post_—modified_—table_—list(2)=NULL;
  • [0085] post_—modified_—table_—list(3)=NULL;
  • [0086] query_—list(1)=‘select msg#, orig_—id, msg_—type from A’;
  • [0087] query_—list(2)=‘select msg#, text from A where msg_—type=1’;
  • [0088] query_—list(3)=‘select msg#, text from A where msg_—type=2’;
    where owner_—list is a collection of schema names and interim_—table_—list is a collection of interim table names.

[0089] After these initializations, the appropriate database procedure may be invoked as follows:

• • [0090] Dbms_—redefiniton.start_—redef_—table(owner_—list, interim_—table_—list, post_—modified_—table_—list, query_—list);

[0091] In one embodiment, when the online modification procedure is invoked, the following events take place:

• • [0092] 1) A “Materialized View” log is created on table A:
  • [0093] 2) Three Materialized Views are created:

[0094] For example, a materialized view Temp_—A is created using prebuilt table as select msg#, orig_—id, msg_—type from A; a materialized view A2 is created using prebuilt table as select msg#, text from A where msg_—type=1; and a materialized view A3 is created using prebuilt table as select msg#, text from A where msg_—type=2.

[0095] The creation of the materialized views can take up to several hours. However, during this time, users still have access to the target table A.

[0096] The next step is to synchronize the data of the target table A with the revised tables, Temp_—A, A2, A3 by calling the dbms_—redefinition.sync_—interim_—table procedure. This step refreshes the materialized views. According to one embodiment, the following command may be used to perform the synchronization:

• • [0097] DBMS_—REDEFINITION.SYNC_—REDEF_—TABLE(owner_—list, interim_—table_—list)

[0098] Finally, the online modification is completed by calling the dbms_—finish_—redef table procedure. According to one embodiment, the call has the following form:

• • [0099] DBMS_—REDEFINTION.FINSH_—REDEF_—TABLE(owner_—list, interim_—table list, post_—modified_—table_—list)

[0100] In this step, the following occurs:

• • [0101] 1. Materialized Views of Temp_—A, A2, A3 are refreshed. The time for the refresh may be on the order of minutes.
  • [0102] 2. Table A is locked at which time no more updates to the data in Table A are allowed. This is where the downtime begins. At this time, the 2001 version of the application should not be used any more.
  • [0103] 3. After the table A is locked, the materialized views of Temp_—A, A2, A3 are refreshed again.
  • [0104] 4. Table A's identity is switched with table Temp_—A. Neither table A2 nor A3 needs its identity switched.
  • [0105] 5. Table A (which used to be Temp_—A) is unlocked. It was probably unavailable for only a period of seconds.

[0106] At this time, the 2002 version of the application can be used. Downtime was on the order of seconds instead of hours by using online redefinition. The small amount downtime is typically not apparent to the user who is trying to access the target table.

[0107] Assume that in the year 2003, the above database application has changed again to require modification of tables in the relational database system. Assume that performance of the application went down due to the conversion of table A to the 3rd normal form because of the continuous joins between A, A2 and A3. As a result, the 2003 version denormalizes the tables from tables A, A2, A3 back into the 2001 version of table A.

[0108] The first step for modifying tables A, A2, and A3 into the 2001 version of table A is to create a revised table as follows:

• • [0109] create table Temp_—A (msg# number, orig_—id number, msg_—type number, msg_—text varchar2(4000));

[0110] Next, a query list is created as follows:

• • [0111] Table Temp_—A is defined as: select msg#, orig_—id, msg_—type, address from A, A2 where A.msg#=A2.msg# union select msg#, orig_—id, msg_—type, comment from A, A3 where A.msg#=A3.msg#;

[0112] The list of post-modified tables comprises Table A. The list of “revised tables” are Temp_—A. Assume that all tables are owned by user U.

[0113] The online modification may be started by performing the following initializations:

• • [0114] owner_—list(1):=‘U’
  • [0115] interim_—table_—list(1):=‘Temp_—A’
  • [0116] post_—modified_—table_—list(1):=‘A’;
  • [0117] query_—list(1):=‘select msg#, orig_—id, msg_—type, address from A, A2 where A.msg#=A2.msg# union select msg#, orig_—id, msg_—type, comment from A, A3 where A.msg#=A3.msg#’
    where exist_—table_—list is a collection of existing tables and new_—table_—list is a collection of new tables.

[0118] After the initializations have been performed, the appropriate database procedure may be invoked as follows:

• • [0119] dbms_—redefinition.start_—redef table (owner_—list, interim_—table_—list, post_—modified_—table_—list, query_—list);

[0120] In one embodiment, when the online modification procedure is invoked, the following events take place:

• • [0121] 1. A Materialized View log is created on table A, A2, A3.
  • [0122] 2. One Materialized View is created. For example, a materialized view Temp_—A using prebuilt table as select msg#, orig_—id, msg_—type, address from A, A2 where A.msg#=A2.msg# union select msg#, orig_—id, msg_—type, address from A, A3 where A.msg#=A3.msg#;

[0123] The creation of the materialized views can take up to several hours. However, during this time, users still have access to the target table A. The next step is to synchronize the data of the target tables A, A2 and A3 with the revised table, Temp_—A. This synchronization may be performed through the following call:

• • [0124] DBMS_—REDEFINITION.SYNC_—REDEF_—TABLE(owner_—list, interim_—table_—list)

[0125] Finally, the online modification is completed by calling the dbms_—finish_—redef table procedure. This may be performed by the following call:

• • [0126] DBMS_—REDEFINITION.FINISH_—REDEF_—TABLE(owner_—list, interim_—table-list, post_—modified_—table_—list)

[0127] In this step, the following occurs:

• • [0128] 1. Materialized View Temp_—A is refreshed. The time for the refresh may be on the order of minutes.
  • [0129] 2. Tables A, A2 and A3 are locked. This is where the downtime begins. At this time, the 2002 version of the application may not be used any more.
  • [0130] 3. Table A's identity is switched with table Temp_—A.
  • [0131] 4. After tables A, A2 and A3 are locked, the materialized view Temp_—A is refreshed again.
  • [0132] 5. Table A (which used to be Temp_—A) is unlocked. It was probably unavailable for only a period of seconds.

[0133] At this time, the 2003 version of the application can be used. Downtime was on the order of seconds instead of hours by using online redefinition.

[0134] A table can be subdivided into subdivisions called “partitions”.

[0135] For this example consider two non-partitioned tables CUSTOMER(cust_—id, name, address) and orders (order_—id, customer_—id, order_—line_—id, item_—id, item_—price) which need to be re-defined online as follows:

[0136] The customer table will have a new column which will contain the sum of all orders for that customer.

[0137] The orders table will be split into two tables. The first table which will be named orders after the redefinition will contain the order_—id, customer_—id and the total for this individual order. The second table which will be named order_—line after the redefinition will contain the orderline_—id, order_—id, item_—id and the item_—price.

[0138] The post-redefinition order_—line will be partitioned by range on order_—id.

[0139] In this case the user will have to create the three interim tables as follows.

• • [0140] create table int_—customer
  • [0141] (cust_—id number,
  • [0142] name varchar2 (30),
  • [0143] address varchar2 (1000),
  • [0144] cust_—totals number);
  • [0145] create table int_—orders
  • [0146] (order_—id number,
  • [0147] cust_—id number,
  • [0148] order_—totals number);
  • [0149] create table order_—line
  • [0150] (order_—line_—id number
  • [0151] order_—id number,
  • [0152] item_—id number,
  • [0153] item_—price number)
  • [0154] partition by range (order_—id)
  • [0155] (partition ord1000 values less than (1000) tablespace tbs_—1,
  • [0156] partition ord2000 values less than (2000) tablespace tbs_—1,
  • [0157] partition ord2000 values less than (2000) tablespace tbs_—1);

[0158] Then the user will start the redefinition process by executing the following call:

• • [0159] uname_—list(1):=‘U1’;
  • [0160] uname_—list(2):=‘U2’;
  • [0161] uname_—list(3):=‘U3’;
    • [0162] list of interim tables
      • [0163] int_—table_—list(1):=‘INT_—CUSTOMER’;
      • [0164] int_—table_—list(2):=‘INT_—ORDERS’;
      • [0165] int_—table_—list(3):=‘ORDER_—LINE’;
    • [0166] corresponding list of post-redef table names (one for each interim table)
      • [0167] post_—table_—list(1):=‘CUSTOMER’;
      • [0168] post_—table_—list(2):=‘ORDERS’
      • [0169] post_—table_—list(3):=”;

[0170]

list of corresponding mapping queries (one for each interim table)
query—list(1) := ‘select c.cust—id, c.name, c.address,
	sum(o.item—price) cust—totals
	from customer c, orders o
	where c.cust—id = o. cust—id
	group by c.cust—id’;
query—list(2) := ‘select o.order—id, o.cust—id,
	sum(o.item—price) order—totals
	from orders o
	group by o.order—id’;
query—list(3) := ‘select order—line—id, order—id, item—price from orders’;
dbms—redefinition.start—redef—table(uname—list,
	int—table—list,
	post—table—list,
	query—list;

[0171] During this step, the following occur:

• • [0172] 1. Materialized view logs on CUSTOMERS and ORDERS are created;
  • [0173] 2. Materialized view INT_—CUSTOMER, INT_—ORDERS and ORDER_—line are created using the corresponding queries in the query_—list.

[0174] After this step the user can then create any triggers, indices and constraints on the interim tables. Any referential constraints involving these interim tables should be created disabled (they will be automatically enabled after the redefinition). The user should also define any grants associated with the interim tables and these grants will replace the grants on the original table after redefinition.

[0175] The user can optionally synchronize the interim tables by calling sync_—interim_—table( )

• • [0176] dbms_—redefinition.sync_—redef_—table(uname_—list, int_—table_—list);

[0177] The user can then complete the redefinition process by calling

[0178]

	finish—redef—table( )
	dbms—redefinition.finish—redef—table(uname—list,
	int—table—list);
	post—table—list);

[0179] During this step, the following occur:

• • [0180] 1. The materialized views INT_—CUSTOMER, INT_—ORDERS and ORDER_—LIN are refreshed;
  • [0181] 2. The target tables CUSTOMER and ORDERS are locked.
  • [0182] 3. After the target tables CUSTOMER and ORDERS have been locked, the materialized view INT_—CUSTOMER, INT_—ORDERS and ORDER_—LIN are refreshed once again.
  • [0183] 4. The identities of the table CUSTOMER and INT_—CUSTOMER are switched. Similarly, the identities of the table ORDERS and INT_—ORDERS are switched.
  • [0184] 5. The materialized views INT_—CUSTOMER, INT_—ORDERS and ORDER_—LINE and the materialized view logs on CUSTOMER and ORDERS are dropped.
  • [0185] 6. The tables CUSTOMER and ORDERS are unlocked.

[0186] A table can be subdivided into tables with subdivisions called “partitions”. The techniques described herein with reference to “one table to one table”, “one table to many tables”, “many tables to one table”, “many tables to many tables” can be applied to partitions of a table.

[0187] Consider a table T1 with two partitions p1 and p2. In this case, the user wants to online redefine partition p1 only. The user will first create an interim table int_—p1 which will eventually replace the partition p1. The logical structure of the interim table must be consistent with that of the original partition.

[0188] The following are steps performed by the user:

• • [0189] create interim table int_—p1 (post-redefinition version of current partition p1)
  • [0190] start the redefinition process by calling start_—redef_—fragment( )
    • [0191] dbms_—redefinition.start_—reder_—fragment (‘U1’, ‘T1’, ‘P1’, ‘INT_—P1’)

[0192] The user can optionally synchronize the interim table using dbms_—redefinition.sync_—interim_—fragment (‘U1’, ‘T1’, ‘INT_—1’);

[0193] The user can complete the redefinition process by calling

• [0194] finish_—redef_—fragment ( )
  • [0195] dbms_—redefinition.finish_—redef_—fragment (‘U1’, ‘T1’, ‘P1’, ‘INT_—1’);

[0196] FIG. 2 is a block diagram that illustrates a computer system 200 upon which an embodiment of the invention may be implemented. Computer system 200 includes a bus 202 or other communication mechanism for communicating information, and a processor 204 coupled with bus 202 for processing information. Computer system 200 also includes a main memory 206, such as a random access memory (RAM) or other dynamic storage device, coupled to bus 202 for storing information and instructions to be executed by processor 204. Main memory 206 also may be used for storing temporary variables or other intermediate information during execution of instructions to be executed by processor 204. Computer system 200 further includes a read only memory (ROM) 208 or other static storage device coupled to bus 202 for storing static information and instructions for processor 204. A storage device 210, such as a magnetic disk or optical disk, is provided and coupled to bus 202 for storing information and instructions.

[0197] Computer system 200 may be coupled via bus 202 to a display 212, such as a cathode ray tube (CRT), for displaying information to a computer user. An input device 214, including alphanumeric and other keys, is coupled to bus 202 for communicating information and command selections to processor 204. Another type of user input device is cursor control 216, such as a mouse, a trackball, or cursor direction keys for communicating direction information and command selections to processor 204 and for controlling cursor movement on display 212. This input device typically has two degrees of freedom in two axes, a first axis (e.g., x) and a second axis (e.g., y), that allows the device to specify positions in a plane.

[0198] The invention is related to the use of computer system 200 for implementing the techniques described herein. According to one embodiment of the invention, those techniques are implemented by computer system 200 in response to processor 204 executing one or more sequences of one or more instructions contained in main memory 206. Such instructions may be read into main memory 206 from another computer-readable medium, such as storage device 210. Execution of the sequences of instructions contained in main memory 206 causes processor 204 to perform the process steps described herein. One or more processors in a multi-processing arrangement may also be employed to execute the sequences of instructions contained in main memory 206. In alternative embodiments, hard-wired circuitry may be used in place of or in combination with software instructions to implement the invention. Thus, embodiments of the invention are not limited to any specific combination of hardware circuitry and software.

[0199] The term “computer-readable medium” as used herein refers to any medium that participates in providing instructions to processor 204 for execution. Such a medium may take many forms, including but not limited to, non-volatile media, volatile media, and transmission media. Non-volatile media includes, for example, optical or magnetic disks, such as storage device 210. Volatile media includes dynamic memory, such as main memory 206. Transmission media includes coaxial cables, copper wire and fiber optics, including the wires that comprise bus 202. Transmission media can also take the form of acoustic or light waves, such as those generated during radio wave and infrared data communications.

[0200] Common forms of computer-readable media include, for example, a floppy disk, a flexible disk, hard disk, magnetic tape, or any other magnetic medium, a CD-ROM, any other optical medium, punch cards, paper tape, any other physical medium with patterns of holes, a RAM, a PROM, and EPROM, a FLASH-EPROM, any other memory chip or cartridge, a carrier wave as described hereinafter, or any other medium from which a computer can read.

[0201] Various forms of computer readable media may be involved in carrying one or more sequences of one or more instructions to processor 204 for execution. For example, the instructions may initially be carried on a magnetic disk of a remote computer. The remote computer can load the instructions into its dynamic memory and send the instructions over a telephone line using a modem. A modem local to computer system 200 can receive the data on the telephone line and use an infrared transmitter to convert the data to an infrared signal. An infrared detector coupled to bus 202 can receive the data carried in the infrared signal and place the data on bus 202. Bus 202 carries the data to main memory 206, from which processor 204 retrieves and executes the instructions. The instructions received by main memory 206 may optionally be stored on storage device 210 either before or after execution by processor 204.

[0202] Computer system 200 also includes a communication interface 218 coupled to bus 202. Communication interface 218 provides a two-way data communication coupling to a network link 220 that is connected to a local network 222. For example, communication interface 218 may be an integrated services digital network (ISDN) card or a modem to provide a data communication connection to a corresponding type of telephone line. As another example, communication interface 218 may be a local area network (LAN) card to provide a data communication connection to a compatible LAN. Wireless links may also be implemented. In any such implementation, communication interface 218 sends and receives electrical, electromagnetic or optical signals that carry digital data streams representing various types of information.

[0203] Network link 220 typically provides data communication through one or more networks to other data devices. For example, network link 220 may provide a connection through local network 222 to a host computer 224 or to data equipment operated by an Internet Service Provider (ISP) 226. ISP 226 in turn provides data communication services through the worldwide packet data communication network now commonly referred to as the “Internet” 228. Local network 222 and Internet 228 both use electrical, electromagnetic or optical signals that carry digital data streams. The signals through the various networks and the signals on network link 220 and through communication interface 218, which carry the digital data to and from computer system 200, are exemplary forms of carrier waves transporting the information.

[0204] Computer system 200 can send messages and receive data, including program code, through the network(s), network link 220 and communication interface 218. In the Internet example, a server 230 might transmit a requested code for an application program through Internet 228, ISP 226, local network 222 and communication interface 218. In accordance with the invention, one such downloaded application implements the techniques described herein.

[0205] The received code may be executed by processor 204 as it is received, and/or stored in storage device 210, or other non-volatile storage for later execution. In this manner, computer system 200 may obtain application code in the form of a carrier wave.

[0206] In the foregoing specification, the invention has been described with reference to specific embodiments thereof. It will, however, be evident that various modifications and changes may be made thereto without departing from the broader spirit and scope of the invention. The specification and drawings are, accordingly, to be regarded in an illustrative rather than a restrictive sense.