Error detection/correction and fault detection/recovery – Data processing system error or fault handling – Reliability and availability
Reexamination Certificate
1998-09-28
2001-05-01
Ray, Gopal C. (Department: 2181)
Error detection/correction and fault detection/recovery
Data processing system error or fault handling
Reliability and availability
C707S793000, C711S162000
Reexamination Certificate
active
06226759
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to digital data backup systems. More particularly, the invention concerns a technique for performing an expedited data backup by creating a duplicate set of pointers to a current dataset already identified by an original pointer set, then designating the dataset as a backup dataset, and thereafter preventing changes to the pointed-to-data and the duplicate pointers, where changes to the current dataset are nonetheless effected by storing new data and modifying the original pointer set alone.
2. Description of the Related Art
With the increasing tide of digital information today, computer users encounter more data than ever to transmit, receive, and process. Data transmission and receipt speeds are continually increasing with each new advance in modems, fiber optics, ISDN, cable television, and other technology. Processing speeds are similarly advancing, as evidenced by the frequent introduction new products by the microprocessor industry.
In addition to transmitting, receiving, and processing data, storing data is another critical need for many users. In fact, many users demand high performance data storage systems to contain huge amounts of data, and to quickly access the data. Engineers are constantly making significant improvements in their storage systems by reducing storage density and increasing storage speed.
For many businesses, data storage is such as critical function that data loss cannot be tolerated. As a result, different techniques and systems for data backup have become widespread. Some examples include the peer-to-peer remote copy system (“PPRC”) and extended remote copy system (“XRC”), both developed by International Business Machines Corp. (“IBM”).
In many applications, it is not only essential to have backup data, but to have quick recovery from backup data in the event of data failure. Some applications that rely on the ability to quickly access stored data include automated teller networks of banks, financial information of stock brokers, reservation systems of airlines, and the like. In applications such as these, slow recovery from failed data can mean lost revenue.
Data backup/recovery occurs in various contexts including an “on-line” environment and an “off-line” environment. In the on-line environment, stored data is continually available to users, and backup operations must therefore be conducted in the “background.” This can slow users' access to their data, including operations such as storing new data, updating existing data, and retrieving stored data. From the user's perspective, slower data access is a disadvantage because it causes frustration and lengthens the time needed to complete projects that require data access. From the system manufacturer's perspective, slower data access is a disadvantage because it makes the storage system less competitive with other manufacturer's storage systems.
Aside from the backup completion time, another concern in the on-line environment is the time needed to recover from backup data when the original data fails. When stored data does fail, it is important to restore the data from the backup copy as quickly as possible. From the user's perspective, data recovery time is part of the data access time, which should be as brief as possible.
In the off-line context, data storage jobs are consolidated for more efficient processing together during a batch processing “window.” The storage system usually goes “off-line” during the batch processing window, and is therefore unavailable to serve requests other than the pre-consolidated jobs being processed. Data backups may be performed regularly during the batch processing window, in serial fashion with the other jobs underway. Consequently, the data backups increase the overall size of the batch processing window, therefore lengthening the time that the system is unavailable to users.
Any off-line recovery that is required during the batch processing window similarly lengthens the time that the storage system is unavailable to users.
As shown above, a number of different backup systems already exist, and certain of these systems constitute significant advances and even enjoy widespread commercial success today. Nonetheless, IBM continually works to improve the performance and efficiency of data backup systems. Some areas of particular focus include minimizing the backup and recovery times in the on-line and off-line storage environments.
SUMMARY OF THE INVENTION
Broadly, the present invention concerns a technique for performing an expedited data backup. This is done by creating a duplicate set of pointers to a current dataset already identified by an original pointer set, then designating the dataset as a backup dataset, and thereafter preventing changes to the pointed-to-data and the duplicate pointers, where changes to the current dataset are nonetheless effected by storing new data and modifying the original pointer set alone.
More specifically, the backup involves an original dataset contained in a storage system. This dataset has multiple subparts residing at various original storage locations. The storage system uses a primary set of pointers to identify the original storage locations. The backup is started by duplicating the primary set of pointers to create a backup set of pointers. The backup set of pointers are entered in a backup list, in association with the backup dataset.
To preserve the backup data, the system prevents changes to the original dataset and the backup set of pointers. In response to receiving any changes to the original dataset, the system creates a changed dataset by storing the changes at one or more change storage locations and updating the primary set of pointers to identify the change storage locations and any original storage locations containing unchanged subparts of the original dataset.
To minimize the loss of data in the event of a failure, further backups are performed using similar techniques. Specifically, the modified original set of pointers are duplicated to create a second backup set of pointers. The system enters the second backup set of pointers in the backup list, in association with the backup dataset: Further changes to the changed dataset are prevented. In response to receipt of any further changes to the changed dataset, the system stores the further changes at one or more further changed locations and modifies the modified original set of pointers to identify the further changed storage locations and any original storage locations containing unchanged subparts of the original dataset.
In addition to expediting data backups, the invention speeds data recovery as well. In response to loss of a dataset, the system accesses the backup list and selects one of the listed backup datasets. Then, the selected backup is used to recover the lost dataset. This is rapidly achieved by updating the primary set of pointers to replicate the backup pointer set.
Accordingly, as shown above, one embodiment of the invention may be implemented to provide a method to perform an immediate data backup by duplicating pointers and freezing pointer/data counterparts. In another embodiment, the invention may be implemented to provide an apparatus, such as a data storage system, configured to perform an immediate data backup by duplicating pointers and freezing pointer/data counterparts. In still another embodiment, the invention may be implemented to provide a signal-bearing medium tangibly embodying a program of machine-readable instructions executable by a digital data processing apparatus to perform an immediate data backup by duplicating pointers and freezing pointer/data counterparts.
The invention affords its users with a number of distinct advantages. First, the invention expedites data backup operations. This minimizes system downtime, user waiting, and other undesirable effects of slow backup operations. Data recovery is likewise accelerated with this invention. The invention also provides a number of other advantages and ben
Miller Donald Lee
Sovik Mark Anthony
Dan Hubert & Assoc.
International Business Machines - Corporation
Ray Gopal C.
LandOfFree
Method and apparatus for immediate data backup by... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for immediate data backup by..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for immediate data backup by... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2466991