Method of identifying DB2 objects to be recovered from an unavailable DASD volume without need to access the DASD volume itself and without need for pre-failure preparation

• US 5,625,817 A
• Filed: 02/28/1995
• Issued: 04/29/1997
• Est. Priority Date: 02/28/1995
• Status: Expired due to Term

• Current Assignee: BMC Software Incorporated (KKR & Co., Inc.)
• Original Assignee: BMC Software Incorporated (KKR & Co., Inc.)
• Inventors: Wood, Greg D., Birk, Carla C.
• Primary Examiner(s): Black, Thomas G.
• Assistant Examiner(s): LEWIS, CHERYL RENEA
• Application Number:
  US08/395,992
• Time in Patent Office: 791 Days
• Field of Search: 395/600, 395/180, 395/575, 341/51
• US Class Current: 1/1
• CPC Class Codes: G06F 11/1469 Backup restoration techniques
• CPC Class Codes: G06F 2201/80 Database-specific techniques
• CPC Class Codes: Y10S 707/99942 Manipulating data structure, e.g. compression, compaction, compilation

Abstract

A computer performing a method in accordance with the invention receives from an end-user identifier(s) for one or more damaged (or otherwise unavailable) DASD volumes, and, optionally, an identifier for one or more DB2 subsystems. By searching and correlating DB2 and MVS data resources not residing on the damaged volume(s), the computer determines which object(s) on the specified damaged volume(s) are DB2 objects, and which DB2 subsystem(s) own each of these DB2 objects. A list of recoverable DB2 objects is created, and, optionally, provided to a recovery utility program for the purpose of facilitating recovery. The method permits identification of recoverable DB2 objects without needing access to the damaged or unavailable volumes, either before or after the volume failure, thereby advantageously enabling recovery methods after a volume failure has occurred.

Claims

1. A machine-executed method of identifying one or more DB2 database objects stored on an unavailable DASD volume and associated with a specified DB2 subsystem, wherein information about said one or more DB2 database objects is stored in an MVS catalog structure storage and in a DB2 catalog structure storage, said method comprising:

(a) searching said MVS catalog structure storage and said DB2 catalog structure storage for one or more descriptors that conform to a DB2 naming convention for physical DB2 datasets, each said descriptor corresponding to a logical DB2 object and being referred to as a fully-qualified descriptor, thereby producing a list of said fully-qualified descriptors; and

(b) successively filtering said list of fully-qualified descriptors to create, in a storage, a list of identifiers corresponding to logical DB2 objects associated with said specified DB2 subsystem, said list of identifiers being referred to as a recovery list.

2. A method in accordance with claim 1 and further comprising:

(c) executing a recovery utility program to recover each said logical DB2 object corresponding to a respective said identifier in said recovery list.

3. A method in accordance with claim 1 and further comprising:

(c) providing said recovery list to a recovery utility program to recover each said logical DB2 object corresponding to a respective said identifier in said recovery list.

4. A method in accordance with claim 1, wherein said searching of said MVS catalog structure storage is performed by obtaining a VCAT name for a DB2 directory, said DB2 directory being one of said logical DB2 objects, and by searching an active MVS ZPARMS control block for said one or more fully-qualified descriptors, and using said VCAT name to construct a search-target prefix to be used in subsequent searching of said MVS catalog structure storage.

5. A method in accordance with claim 1, wherein said searching of said MVS catalog structure storage is performed by obtaining a VCAT name for a DB2 directory, said DB2 directory being one of said logical DB2 objects, using information contained in a user-specified input, and using said VCAT name to construct a search-target prefix to be used in subsequent searching of said MVS catalog structure storage.

6. A method in accordance with claim 1, wherein said searching of said DB2 catalog structure storage is performed by searching a SYSIBM.SYSTABLEPART table and a SYSIBM.SYSINDEXPART table in said DB2 catalog structure storage.

7. A method in accordance with claim 4, wherein said searching of said MVS catalog structure storage is performed by searching said MVS catalog structure for all physical dataset names matching a search-target prefix pattern.

8. A method in accordance with claim 7, wherein said search-target prefix pattern is of the form <

VCAT name>

.DSNDBD.

9. A method in accordance with claim 7, wherein said searching of said MVS catalog structure storage is performed by invoking the IBM IDCAMS utility.

10. A method in accordance with claim 1, wherein said DB2 naming convention is of the form <

VCAT name>

.DSNDBD.<

dbname>

.<

spname>

.I0001.A<

dsnum>

.

14. A program storage device readable by the machine of a specified one of claims 1-13 and encoding a program of instructions for performing the operations recited in said specified claim.

11. A machine-executed method of identifying for recovery one or more DB2 database objects associated with a specified DB2 subsystem, said one or more DB2 database objects stored on an unavailable DASD volume specified for recovery, wherein information pertaining to said one or more DB2 database objects is stored and accessible in an MVS catalog structure storage and in a DB2 catalog structure storage, said method comprising:

(a) searching said MVS catalog structure storage for all physical dataset names matching a search-target prefix;

(b) building a list of all physical dataset names having a volume identifier matching the volume identifier for said unavailable DASD volume;

(c) filtering said list of all physical dataset names to include only those dataset names conforming to a standard DB2 dataset naming convention, each said physical dataset name contained in said list being referred to as a candidate object;

(d) searching said DB2 catalog structure storage to identify the DB2 subsystem owning each said candidate object;

(e) eliminating from said list of all physical dataset names each said candidate object that is not a table space and not an index;

(f) eliminating from said list of all physical dataset names each said candidate object that is determined to be a table space that does not have a corresponding match in the SYSIBM.SYSTABLEPART table;

(g) eliminating from said list of all physical dataset names each said candidate object that is determined to be an index that does not have a corresponding match in the SYSIBM.SYSINDEXPART table; and

(h) eliminating from said list of all physical dataset names any duplicates of each said candidate object that is determined to be an index, the existence of said duplicates indicated by the unavailability of dataset-level recovery of said index.

12. A method in accordance with claim 11 and further comprising:

(i) eliminating from said list of all physical dataset names each said candidate object not extending to said volume for recovery.

13. A method in accordance with claim 12 and further comprising:

(j) converting said list of all physical dataset names to DB2 object identifiers, said DB2 object identifiers being of a form usable by available recovery utilities.