Data processing: database and file management or data structures – Database design – Data structure types
Reexamination Certificate
1999-12-20
2002-12-03
Rones, Charles L. (Department: 2175)
Data processing: database and file management or data structures
Database design
Data structure types
C707S793000, C345S215000
Reexamination Certificate
active
06490598
ABSTRACT:
FIELD OF THE INVENTION
This invention is directed towards data storage systems, and more particularly towards physical backup and restore of databases residing in data storage systems.
BACKGROUND
Computer systems allow the processing of massive quantities of data for a variety of purposes. As the ability to process data has increased, so has the need for data storage systems which provide massive data storage capabilities combined with fast access for host systems. Another feature required by many businesses and industries is continuous availability. Many businesses operate on a world-wide basis, and have a need for round-the-clock access to databases stored in one or more data storage systems. The data stored in these data storage systems is changing at an incredible rate, for example with transaction processing, reservation systems and data mining, the data is changing and updating many times per second.
Another requirement for data storage systems is periodic backup of data both for archival purposes and for data recovery in case of a system failure. For many businesses, a loss of data can be catastrophic. Therefore, system backups must be performed on a frequent basis.
However, the need for system backups often interferes with the need for continuous availability. With many data storage systems, performing a system backup requires taking the data storage system offline, thereby denying continuous access to the data.
One solution to this problem is used for RAID (Redundant Array of Independent Disks) systems. In RAID-1 systems, two physical storage devices, such as disks, each store identical data, in a process known as “mirroring”. This provides a very high level of fault tolerance in the form of redundancy, and it also allows data backups to be performed while still allowing continuous data access. Typically, the mirroring process is stopped (referred to as splitting the mirrors), and one of the disks is taken off-line and backed up, while the other disk remains online and available. When the first is completely backed up, the two are resynchronized (so that the data is identical on both), and the data storage system returns to full operation.
However, there are still problems related to backing up at the physical disk volume level, instead of at the application level. For example, a database (whether hierarchical, relational, object-oriented or otherwise) stores data in a logical structure which does not match physical disk details. Several layers of mapping are performed to map the database data onto the physical disks. Modem data storage systems perform mapping of physical disks to logical volumes, to support a standard representation of storage units to host systems. These logical volumes appear to host systems as a defined set of storage volumes for the host to access. The data storage systems perform all functions of converting logical volume addressing and accessing to functions effective on physical disks.
But there is still one layer of mapping (or more) from the database application to logical volumes. The database application performs much of this mapping function, to provide the host systems with access to the database data in a format dictated by the database application. The host systems benefit from having all the mapping details performed automatically by the database applications and the data storage systems. However, the layers of mapping make the process of backup and restore much more difficult. Traditional backup systems are unable to back up the variety of data formats and client platforms that exist in heterogeneous, growing networks, leaving potential holes in network backup coverage and leaving administration decentralized. There is little integration of the backup systems to allow “intelligent” backups by taking advantage of the mapping layers, for example to perform incremental backups. Often, the only solution is for “brute force” complete database space backups, which are inefficient and time consuming. This causes a great discrepancy between the advantages which many database applications provide to host systems. Ultimately, there is significant difficulty in performing timely data storage system backups as necessary to prevent catastrophic data loss.
Some presently available database applications have a non-integrated approach for external backup of mirrored database spaces. This approach blocks the data storage system server, allowing only read-only accesses to run, therefore it is only usable for a mirrored data storage system as previously described. Further, this approach requires blocking the data storage system server for the length of time necessary to split mirrors. Also, the user must manually perform logical restore to successfully restore their spaces. These non-integrated approaches essentially involve making the DBMS accept on faith that the user has done a physical restore. There is great risk in depending on the users to perform all backup/restore procedures correctly.
SUMMARY
The present invention provides for intelligent, integrated external backup and restore for DBMS which is stored on RAID data storage systems.
According to the present invention, an interface between the data storage system backup system and the DBMS allows the backup system to work with the DBMS for greatly enhanced backup and restore. External backups can be created by blocking the server after forcing a checkpoint, whereupon the backup is created using an external resource. In one embodiment, this is accomplished by disconnecting the mirrored target data from the source online data, so that the target data can be saved to a safer site other than the production site. After the backup is created, the server is unblocked to resume normal server operations.
An example of a relational database is the Informix 7.3 database and database management system (DBMS) from Informix Software Inc. of Menlo Park, Calif. The Informix DBMS runs on several Unix platforms as well as Microsoft Windows®. The Informix DBMS provides very limited access to other applications attempting to perform system utility operations such as backup/restore. An illustrative embodiment of the present invention provides for external backup/restore of Informix 7.3 database spaces and blob spaces, through an interface to the Informix OnLine server.
During an unplanned event such as a data center disaster, fast restores can be performed from the external backups. In an external backup, the SMV (Storage Management Vendor) is completely responsible for issuing all I/O commands to move the data from disk to tape. In a normal DBMS managed backup, the database vendor issues the I/O command to read the data from disk, and the SMV then writes the data to tape. Conversely, in an external restore, the SMV issues all I/O commands from the tape back to disk. A DBMS managed restore has the SMV reading the data from tape, and the DBMS then writing the data to the disk. In both cases, data is restored from tape, and the roll forward (if required) with logical logs may occur.
According to an illustrative embodiment of the present invention, there are two different types of external restore supported: complete external restore and partial external restore. In a complete external restore, all spaces will be restored to the most recent checkpoint that was generated while creating an external backup. A complete external restore is usually necessary when a major portion of the data storage system server needs to be restored or if an application error corrupts data. If users lose only a portion of the data (which is more typically the case), a partial external restore may be performed. A partial external restore restores only a subset of spaces that are down. This subset of spaces is defined by the user, with the limitation that it does not include any critical database spaces.
After a complete external restore, the user can optionally do a logical restore to bring the server to logical consistency. A point-in-time restore can also be done as part of a complete external restore, to roll forward the logs to a specif
Brown Rudnick Berlack & Israels LLP
EMC Corporation
Lowry David D.
Michaelis Brian L.
Rones Charles L.
LandOfFree
System and method for external backup and restore for a... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System and method for external backup and restore for a..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System and method for external backup and restore for a... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2931375