Whether it’s a natural disaster, software corruption, hardware failure, or simple human error, at some point almost all business experience a situation where data will be put at risk. This is where IT disaster recovery services and effective disaster recovery (DR) planning address five key considerations – downtime, data integrity, cost, simplicity and security.
Data disaster recovery plans for safeguarding business data can include tape, disk, cloud technologies, or some combination of two or more systems. There is no universal DR plan that suits all organisations, but the following considerations are intended to lay the groundwork for disaster planning and preparedness.
Time is money: How much time it takes to recover data – how long a business can afford to be out of business – is a high priority.
Recovering terabytes of data from tape involves first identifying which set is needed, requesting a delivery from the offsite storage provider, correlating each tape with a logbook, determining which to load first, then restoring data. This process may take days of working round-the-clock, depending on the amount of data. Recovering from disk storage devices kept on-site reduces time to hours instead of days, but disk itself is not a viable option if the place of business is damaged. Disk is also vulnerable to corruption and accidental erasures.
Traditional data restores are often less than 100 percent successful – some files are simply gone for good. When these files pertain to customers, transactions, or anything else not easily reproduced, lost data becomes lost revenue.
As a data repository, tape can be unreliable. With a full day between backups, inevitably at least twelve hours of data will be lost if not more.
Disk mirroring provides data redundancy. In the event of a disaster, however, all data produced since the last backup will still be lost. Synchronous replication does not fully protect against data loss because if software is corrupted or data is deleted from the main server, or in a virus attack, that problem will be copied to backups, and older versions of files cannot be accessed. As with tape, disk is dependent on fragile hardware.
As a business becomes more reliant on continuous access to stored data, the more data it will store – and the more it must back up. Tapes themselves are relatively cheap, but considering their fallibility; it is not viable to be dependent on a medium that will lose data if it is bent or dropped. This same logic applies to disk as well. As data volumes grow, so do backup costs.
Disaster is troublesome enough – recovery shouldn’t be. Traditional strategies effectively drop IT administrators into a maze of hardware and bookkeeping, whether it involves putting a tape in the drive and waiting or piecing data together from corrupted disks. Mirroring disk to a second site can improve reliability but often the task of finding a second data centre and establishing a reliable link can be daunting and filled with unexpected complexity. Keeping a restore as simple as possible will help reduce stress levels, and speed up recovery times.
Just as primary storage must be protected, so must its backups. Most reputable backup facilities provide superb security, except that tapes are sometimes lost or damaged in transit. An improperly closed door on the back of a truck, and sensitive business data is on the street. With disk, the security of data is entirely dependent on internal security mechanisms. If they are secure, your backups will be secure, except for cases of human error or malicious attack.
It may sound gloomy and stressful however DR planning is crucial for businesses of all sizes and well executed plans will help minimise impact on the business and speed up times to being back up and running.