Method and apparatus for data recovery optimization in a logically partitioned computer system

Claim 3

The method for managing the availability of compiled data of claim 1 , wherein said compiled data comprises a plurality of database indexes.

Claim 4

The method for managing the availability of compiled data of claim 1 , wherein said predetermined recovery time limit is a user specified recovery time limit.

Claim 5

The method for managing the availability of compiled data of claim 1 , wherein said step of automatically storing said second subset of compiled data comprises logging changes to said second subset of compiled data in at least one non-volatile storage device, and wherein said step of automatically storing said fourth subset of compiled data comprises logging changes to said fourth subset of compiled data in at least one non-volatile storage device.

Claim 6

The method for managing the availability of compiled data of claim 1 , wherein: if said step of automatically detecting a configuration change to said first logical partition detects a configuration change which increases the processor resource assignment of said first logical partition, then said step of automatically determining third and fourth discrete subsets determines a fourth subset which is a subset of said second subset; and if said step of automatically detecting a configuration change to said first logical partition detects a configuration change which reduces the processor resource assignment of said first logical partition, then said step of automatically determining third and fourth discrete subsets determines a third subset which is a subset of said first subset.

Claim 7

The method for managing the availability of compiled data of claim 1 , wherein said step of automatically determining first and second discrete subsets of said compiled data comprises defining a plurality of buckets for said compiled data, each bucket corresponding to a respective rebuild time; determining a respective cumulative rebuild time for each bucket, said cumulative rebuild time representing the rebuild time using said first set of resources of all compiled data in the respective bucket and in all buckets corresponding to smaller rebuild times; and selecting the compiled data in all buckets having cumulative rebuild times less than said predetermined recovery time limit for inclusion in said first discrete subset; and wherein said step of automatically determining third and fourth discrete subsets of said compiled data comprises determining a respective cumulative rebuild time for each bucket, said cumulative rebuild time representing the rebuild time using said second set of resources of all compiled data in the respective bucket and in all buckets corresponding to smaller rebuild times; and selecting the compiled data in all buckets having cumulative rebuild times less than said predetermined recovery time limit for inclusion in said third discrete subset.

Claim 8

The method for managing the availability of compiled data of claim 6 , wherein each said bucket comprises at least one sub-bucket, each sub-bucket corresponding to a processor resource; wherein if said second set of resources is greater than first set, then said step of automatically determining third and fourth discrete subsets comprises allocating additional sub-buckets corresponding to the difference between said first set of resources and said second set of resources, and transferring compiled data units from previously existing sub-buckets to said additional sub-buckets; and wherein if said second set of resources is less than said first set, then said step of automatically determining third and fourth discrete subsets comprises removing sub-buckets corresponding the difference between said first set of resources and said second set of resources, and transferring compiled data units from the removed sub-buckets to remaining sub-buckets.

Claim 9

A program product for managing the availability of compiled data in a computer system having a plurality of central processors and a dynamic logical partitioning mechanism, said dynamic logical partitioning mechanism supporting a plurality of defined logical partitions of said computer system, each logical partition having a respective processor resource assignment, wherein each task executing in said computer system is assigned to a respective one of said logical partitions and wherein the definition of a plurality of logical partitions may be dynamically altered, said program product comprising a plurality of processor executable instructions recorded on computer readable media, wherein said instructions, when executed by at least one central processor of said computer system, cause the system to perform the steps of: automatically determining first and second discrete subsets of said compiled data within a first logical partition, wherein said first discrete subset of said compiled data can be rebuilt using a first set of resources assigned to said first logical partition within at least one predetermined recovery time limit; automatically storing said second subset of said compiled data in a form which does not require rebuilding of said second subset; automatically detecting a configuration change to said first logical partition which alters the processor resource assignment of said first logical partition; automatically determining third and fourth discrete subsets of said compiled data within said first logical partition, wherein said third discrete subset of said compiled data can be rebuilt using a second set of resources assigned to said first logical partition within said at least one predetermined recovery time limit, said second set of resources being the set of resources assigned to said first logical partition after said configuration change, said second set of resources being different from said first set of resources; and thereafter automatically storing said fourth subset of compiled data in a form which does not require rebuilding of said fourth subset.

Claim 10

A program product for managing the availability of compiled data in a computer system having a plurality of central processors and a dynamic logical partitioning mechanism, said dynamic logical partitioning mechanism supporting a plurality of defined logical partitions of said computer system, each logical partition having a respective processor resource assignment, wherein each task executing in said computer system is assigned to a respective one of said logical partitions and wherein the definition of a plurality of logical partitions may be dynamically altered, said program product comprising a plurality of processor executable instructions recorded on computer readable media, wherein said instructions, when executed by at least one central processor of said computer system, cause the system to perform the steps of: automatically determining first and second discrete subsets of said compiled data within a first logical partition, wherein said first discrete subset of said compiled data can be rebuilt using a first set of resources assigned to said first logical partition within at least one predetermined recovery time limit; automatically storing said second subset of said compiled data in a form which does not require rebuilding of said second subset; automatically detecting a configuration change to said first logical partition which alters the processor resource assignment of said first logical partition; automatically determining third and fourth discrete subsets of said compiled data within said first logical partition, wherein said third discrete subset of said compiled data can be rebuilt using a second set of resources assigned to said first logical partition within said at least one predetermined recovery time limit, said second set of resources being the set of resources assigned to said first logical partition after said configuration change, said second set of resources being different from said first set of resources; and thereafter automatically storing said fourth subset of compiled data in a form which does not require rebuilding of said fourth subset.

Claim 11

The program product for managing the availability of compiled data of claim 8 , wherein said compiled data comprises a plurality of database indexes.

Claim 12

The program product for managing the availability of compiled data of claim 8 , wherein said predetermined recovery time limit is a user specified recovery time limit.

Claim 13

The program product for managing the availability of compiled data of claim 8 , wherein said step of automatically storing said second subset of compiled data comprises logging changes to said second subset of compiled data in at least one non-volatile storage device, and wherein said step of automatically storing said fourth subset of compiled data comprises logging changes to said fourth subset of compiled data in at least one non-volatile storage device.

Claim 14

The program product for managing the availability of compiled data of claim 8 , wherein: if said step of automatically detecting a configuration change to said first logical partition detects a configuration change which increases the processor resource assignment of said first logical partition, then said step of automatically determining third and fourth discrete subsets determines a fourth subset which is a subset of said second subset; and if said step of automatically detecting a configuration change to said first logical partition detects a configuration change which reduces the processor resource assignment of said first logical partition, then said step of automatically determining third and fourth discrete subsets determines a third subset which is a subset of said first subset.

Claim 15

The program product for managing the availability of compiled data of claim 8 , wherein said step of automatically determining first and second discrete subsets of said compiled data comprises defining a plurality of buckets for said compiled data, each bucket corresponding to a respective rebuild time; determining a respective cumulative rebuild time for each bucket, said cumulative rebuild time representing the rebuild time using said first set of resources of all compiled data in the respective bucket and in all buckets corresponding to smaller rebuild times; and selecting the compiled data in all buckets having cumulative rebuild times less than said predetermined recovery time limit for inclusion in said first discrete subset; and wherein said step of automatically determining third and fourth discrete subsets of said compiled data comprises determining a respective cumulative rebuild time for each bucket, said cumulative rebuild time representing the rebuild time using said second set of resources of all compiled data in the respective bucket and in all buckets corresponding to smaller rebuild times; and selecting the compiled data in all buckets having cumulative rebuild times less than said predetermined recovery time limit for inclusion in said third discrete subset.

Claim 16

The program product for managing the availability of compiled data of claim 13 , wherein each said bucket comprises at least one sub-bucket, each sub-bucket corresponding to a processor resource; wherein if said second set of resources is greater than first set, then said step of automatically determining third and fourth discrete subsets comprises allocating additional sub-buckets corresponding to the difference between said first set of resources and said second set of resources, and transferring compiled data units from previously existing sub-buckets to said additional sub-buckets; and wherein if said second set of resources is less than said first set, then said step of automatically determining third and fourth discrete subsets comprises removing sub-buckets corresponding the difference between said first set of resources and said second set of resources, and transferring compiled data units from the removed sub-buckets to remaining sub-buckets.

Claim 17

A computer system, comprising: a plurality of central processing units; a memory; a logical partitioning mechanism supporting a plurality of defined logical partitions of said computer system, each logical partition having a respective processor resource assignment, wherein each task executing in said computer system is assigned to a respective one of said logical partitions and wherein the definition of said logical partitions may be dynamically altered; a protection utility for compiled data in a first logical partition of said computer system, wherein said protection utility: (a) automatically computes rebuild times of discrete portions of said compiled data according to a first configuration of said first logical partition to determine a first subset of said compiled data to be stored in a form which does not require rebuilding, said first subset being determined to meet a predetermined recovery time limit; (b) automatically detects a configuration change to said first logical partition which affects the time required to rebuild said compiled data, said configuration change configuring said first logical partition according to a second configuration; (c) responsive to detecting said configuration change, computes rebuild times of said discrete portions of said compiled data according to said second configuration of said first logical partition; and (d) automatically determines a second subset of said compiled data to be stored in a form which does not require rebuilding using said rebuild times of said discrete portions of said compiled data according to said second configuration of said first logical partition, said second subset being determined to meet said predetermined recovery time; and (e) automatically migrates compiled data stored in said form which does not require rebuilding from said first subset to said second subset.

Claim 18

A computer system, comprising: a plurality of central processing units; a memory; a logical partitioning mechanism supporting a plurality of defined logical partitions of said computer system, each logical partition having a respective processor resource assignment, wherein each task executing in said computer system is assigned to a respective one of said logical partitions and wherein the definition of said logical partitions may be dynamically altered; a protection utility for compiled data in a first logical partition of said computer system, wherein said protection utility: (a) automatically computes rebuild times of discrete portions of said compiled data according to a first configuration of said first logical partition to determine a first subset of said compiled data to be stored in a form which does not require rebuilding, said first subset being determined to meet a predetermined recovery time limit; (b) automatically detects a configuration change to said first logical partition which affects the time required to rebuild said compiled data, said configuration change configuring said first logical partition according to a second configuration; (c) responsive to detecting said configuration change, computes rebuild times of said discrete portions of said compiled data according to said second configuration of said first logical partition; and (d) automatically determines a second subset of said compiled data to be stored in a form which does not require rebuilding using said rebuild times of said discrete portions of said compiled data according to said second configuration of said first logical partition, said second subset being determined to meet said predetermined recovery time; and (e) automatically migrates compiled data stored in said form which does not require rebuilding from said first subset to said second subset.

Claim 19

The computer system of claim 15 , wherein said compiled data comprises a plurality of database indexes.

Claim 20

The computer system of claim 15 , wherein said form which does not require rebuilding comprises one or more log records stored in a non-volatile storage device.