Object level recovery refers to a method of restoring a database to a prior state by restoring single data objects individually, instead of restoring the entire database from a backup. It is a quick, clean, and efficient form of object recovery, and one that is less likely to result in data corruption or other complications, as well as being less of a threat to the information and access of individual database users. Where direct recovery of an individual object is possible, its use is far less time- and resource-intensive than restoring from a whole database.
What Makes Object Level Recovery Different?
Imagine that you’re in charge of maintaining a warehouse full of filing cabinets. You arrive at your job on a Monday morning, only to discover that there was a fire over the weekend. The files in a particular corner of the warehouse were damaged. It isn’t a large number of files, and as luck would have it the owners of the warehouse keep a copy of every file in a basement underneath the warehouse. Unfortunately, upon entering the basement, you find that the files are simply stacked everywhere, in no particular order.
You don’t have a lot of information to replace, but you’ve no idea where it is in all of the piles of file folders. The first file you pick up might be the first file that you need, but you might be down there for hours before finding the last one. This is comparable to the level of difficulty that many database administrators face when necessary files are accidentally wiped or deleted. Restoring from a backup requires space in the system for the backup to be uncompressed and examined, and it can take a long time to go through and find the exact files which are needed.
How Can This be Made More Efficient?
Object level recovery works by providing the database administrator with a schematic for that basement room. It allows them to go in and find the specific data objects that they need, without having to go through the time and effort of working through the backup of an entire database. Because modern databases are often so enormous, containing unprecedented amounts of information, this represents a significant savings in terms of resources.
Additional resource: 50 Most Advanced University Computer Science Departments 2016
How Does Object Recovery Software Work?
This type of data recovery requires that those objects which are intended, just in case, to be recovered in this fashion, be designated as such beforehand. OLR software works by taking user input in the form of what types of data (including table data) are to be recovered, in advance. Then it records where to find them within the overall backup. As with the previous example, it creates an internal schematic for later referencing, should it prove necessary. Only the types of data specified by the administrator ahead of time will be so referenced; the more you reference, the longer a restorative action will take, and the more memory it will occupy.
More about Individual Data Object Recovery
The technology is still relatively young, but development is ongoing. There are several leading providers for this kind of object-oriented database recovery on the market today. Omics.org offers an open access journal article on one of them, which provides more information about the inner workings of the technology.