You are here

System backups

8 September, 2015 - 17:30

A recurring theme in this textbook has been the idea that value resides in the information and technology-enabled services rather than in the information technology itself. On a very fundamental level, this means that if your information technology (e.g., an application or web server) is damaged or destroyed, the organization can purchase a replacement. However, if the data on that server is destroyed, it may be much more expensive to replace than the technology. Furthermore, the data may irreplaceable. Depending on the value of the data, the data's loss may result in business failure or a prolonged disruption of services provided by non-profit organizations.

The point is that information stored on an organization's automated system can be very valuable and merits organizational investment to ensure that it is safeguarded. Consequently, it is important that organizational managers understand system backup technology to the extent that they can confirm that their organizational data is safe guarded.

Nairobi Fire and the Loss of Irreplaceable Documents

March 2, 2004. “The blaze started at 0200 local time (2300 GMT on Monday), sending sheets of orange flames into the night sky.... Council fire engines were unable to cope on their own and several ran out of water as no working hydrants could be found near the city hall.... Among the thousands of documents destroyed were maps showing the routes of new road bypasses (BBC, 2004). (http://news.bbc.co.uk/2/hi/africa/3524855.stm). Computers as well as paper may be lost in fires; both can constitute invaluable information resources of the enterprise. While insurance claims may help to restore the building, important documents, including contractual records and city planning drawings, were lost forever in the Nairobi City Council fire. These types of losses can be prevented by having having tested data recovery capabilities in place.

While the technical details of designing and maintaining a robust data storage infrastructure can be daunting, the fundamental principles for implementing storage backups are relatively simple. First among these is developing an understanding that there are different types of system backups. It is not sufficient to learn from your IT support staff that backups are being performed; managers need to ensure that the required types of backups are being performed. Different guides will approach the issue of backup types in varying levels of details.

There are different types of data or information stored on information systems. We discuss just two categories while acknowledging that backup planners will often use a more detailed set of categories. The distinction that we wish to emphasize is between the software programs (we include the operating system and business application software in this category) and the actual data or information that is manipulated by the software. The rationale for this distinction is that one generally expects the operating system and programs to remain fairly stable, i.e., not change too much. The data, on the other hand, likely changes, or grows, daily. An organization does not want to repeatedly backup data that are not changing. This would pose an unnecessary expense. Plus, an operating system or application is lost, it still should be possible to rebuild the system by reloading the required operating system and applications. However, if operational data is lost, it is lost for good, if no backup exists.

We do not mean to imply that organizations are not concerned with backing up their software programs. They are. If the software programs and data have been backed up, an organization will be able to more quickly restore service than if it had to rebuild the system by reloading and reconfiguring the entire operating system and applications programs. Consequently, organizations do need to backup their entire systems, but not necessarily as often as they need to backup their data.

Accordingly, there are two major types of system backups (actually there are more; if interested see Wikipedia article on backups). There are full system backups, where all the software and data residing on the system are backed up. Then there are incremental backups, where only the data that has changed since the last full backup or last incremental backup is saved. The advantage of this approach is to reduce the amount of time required to accomplish the backup. This is a particularly important issue on systems which must be taken out of operational status while performing the backup. Restoration is accomplished by restoring the last full backup and then applying the required incremental backups.

Another backup technique which has proven useful is the use of multiple copies of backups. Maintaining multiple backup copies provides several benefits. First, if data becomes corrupted and is accidentally backed up, it is possible to overwrite the good data. By maintaining multiple copies of backups, it may be possible to detect the data corruption before all versions have been overwritten, allowing recovery from an uncorrupted backup. Second, backup media occasionally fails. If there is only one backup, the restoration fails and the organization is out of luck. The probability of three or more sets of media failing simultaneously is extremely small. Consequently, even if some data is lost due to the need to restore from an older backup, the organization should still be able to recover the majority of its data. Finally, it is common to keep the most recent backup in general proximity to the devices on which it normally resides. The availability of the backup minimizes restoration time in the event of common failures. However, some failures are catastrophic and destroying not only the IT equipment, but the entire facility in which the equipment resides, e.g., flood or fire. In such cases the backup data is destroyed with the equipment and recovery becomes impossible, even when the equipment and facilities are replaced. Consequently, best practices dictates that at least one version of the backups be maintained off site to preclude catastrophic loss of all of the organization's data.

There are a variety of technologies that can be used to support the system backup function. They vary in terms of price, capacity [the amount of data that can be backed up], performance [the speed at which data can be backed up], reliability [the probability that the backup technology or media will fail] and overall functionality [special features included make the backup process easier to perform]. For our purposes, it is only necessary to recognize that the price of the backup escalates with speed and capacity. Thus, organizational management needs to understand the value of its information and information services and the consequences of system and data service interruption or loss so that it can make appropriate investment decisions in the purchase of backup system hardware and software.

There is one last critical principle that management truly needs to understand in it evaluation of its system backup process. That is, the organization does not truly have a backup plan if it is not willing to invest the time and resources to test its data restoration capabilities. More than a few organizations having made the investment in backup technology and have experienced the unpleasant surprise learning that their backup process was faulty. The takeaway is to insist that recovery processes be practiced regularly and particularly when significant system changes are implemented. A backup system that has not been tested should not be considered a backup system.