System and method for switching a data partition

A computer-readable storage medium having computer-executable instructions for performing a method of changing a portion of a table, the method comprising:

• separating a portion of a first table from the first table by switching metadata pointers of the portion of the first table with metadata pointers of a portion in a second table, wherein the separated portion is considered off line and therefore not accessible through the first table when the separated portion is part of the second table;
• making changes to the separated portion to generate a changed portion;
• switching the changed portion back into the first table by switching metadata pointers of the portion of the first table with metadata pointers of the portion of the second table, wherein the changed portion is mapped back into the first table and wherein no data is moved as the changed portion is rejoined with the first table,
• the metadata pointers of the first set being set forth in a first index associated with the database table, the metadata pointers of the second set being set forth in a second index associated with a switching table, the separating as performed by switching metadata pointers of the first set and the second set comprising identifying the second index associated with the switching table from among a plurality of indexes already associated with the switching table and switching the first index and the identified second index without modifying the identified second index, the identifying of the second index from among the plurality of indexes comprising finding an index from among the plurality of indexes, the found list having a list of key columns identical to a list of key columns of the first index such that an order and number of key columns matches, the found list also having a set operator of non-key columns identical to a set operator of non-key columns of the first index such that a number and type of the non-key columns matches.

A system for making changes to a first portion of a database table while retaining access to a second portion of the database table, the system comprising:

• a processor having access to memory, the memory having instructions of software components;
• a storage system having the first portion of a database and the second portion of the database;
• a software component having instructions which, when executed by the processor, separates the first portion of the database table from the database table, wherein the separated first portion is accessed off-line from the database table, wherein the first portion of the database table has a first set of metadata pointers and a portion outside of the database table has a second set of metadata pointers, wherein the separation is performed by switching metadata pointers of the first and second set;
• wherein the software component is accessed by a user to make changes to the separated first portion to generate a changed first portion and thereafter switches the changed first portion back into the database table by switching the metadata pointers of the first and second set, wherein the changed first portion is mapped back into the database table, wherein no data is moved as the changed first portion is rejoined with the database table, and wherein the second portion of the database table remained accessible to users during a time when the separated first portion was off-line,
• the metadata pointers of the first set being set forth in a first index associated with the database table, the metadata pointers of the second set being set forth in a second index associated with a switching table, the separating as performed by switching metadata pointers of the first set and the second set comprising identifying the second index associated with the switching table from among a plurality of indexes already associated with the switching table and switching the first index and the identified second index without modifying the identified second index, the identifying of the second index from among the plurality of indexes comprising finding an index from among the plurality of indexes, the found list having a list of key columns identical to a list of key columns of the first index such that an order and number of key columns matches, the found list also having a set operator of non-key columns identical to a set operator of non-key columns of the first index such that a number and type of the non-key columns matches.

A processor-implemented method of simultaneously archiving a first portion of a database and adding a new third portion of a database, while retaining access to a second portion of the database, the method comprising:

• granting, via the processor, users access to a first portion and a second portion of a database, wherein the first portion of the database has a first set of metadata pointers, wherein the second portion of the database has a second set of metadata pointers;
• generating the new third portion, wherein the new third portion of the database has a third set of metadata pointers;
• locking user access to the first portion while maintaining user access to the second portion of the database;
• switching the new third portion into the database and simultaneously switching out the first portion by switching the metadata pointers of the first set with the third set, wherein no data is moved as the new third table is mapped into the database and the first portion is mapped out of the database, and wherein the second portion of the database remained accessible to the users and the new third portion of the database is now also accessible to the users,
• the metadata pointers of the first set being set forth in a first index associated with the database table, the metadata pointers of the second set being set forth in a second index associated with a switching table, the separating as performed by switching metadata pointers of the first set and the second set comprising identifying the second index associated with the switching table from among a plurality of indexes already associated with the switching table and switching the first index and the identified second index without modifying the identified second index, the identifying of the second index from among the plurality of indexes comprising finding an index from among the plurality of indexes the found list having a list of key columns identical to a list of key columns of the first index such that an order and number of key columns matches, the found list also having a set operator of non-key columns identical to a set operator of non-key columns of the first index such that a number and type of the non-key columns matches.