2 cissp ® Official Study Guide Eighth Edition

 Backup Best Practices 
No matter what the backup solution, media, or method, you must address several common 
issues with backups. For instance, backup and restoration activities can be bulky and slow. 
Such data movement can signifi cantly affect the performance of a network, especially 
during regular production hours. Thus, backups should be scheduled during the low peak 
periods (for example, at night). 
The amount of backup data increases over time. This causes the backup (and restora-
tion) processes to take longer each time and to consume more space on the backup media. 
Thus, you need to build suffi cient capacity to handle a reasonable amount of growth over 
a reasonable amount of time into your backup solution. What is reasonable all depends on 
your environment and budget. 
With periodic backups (that is, backups that are run every 24 hours), there is always the 
potential for data loss up to the length of the period. Murphy’s law dictates that a server 
never crashes immediately after a successful backup. Instead, it is always just before the 
next backup begins. To avoid the problem with periods, you need to deploy some form of 
real-time continuous backup, such as RAID, clustering, or server mirroring. 
Finally, remember to test your organization’s recovery processes. Organizations often 
rely on the fact that their backup software reports a successful backup and fail to attempt 
recovery until it’s too late to detect a problem. This is one of the biggest causes of backup 
failures.
 Tape Rotation 
There are several commonly used tape rotation strategies for backups: the Grandfather-
Father-Son (GFS) strategy, the Tower of Hanoi strategy, and the Six Cartridge Weekly 

---

Recovery Plan Development 
833
Backup strategy. These strategies can be fairly complex, especially with large tape sets. 
They can be implemented manually using a pencil and a calendar or automatically by using 
either commercial backup software or a fully automated hierarchical storage management 
(HSM) system. An HSM system is an automated robotic backup jukebox consisting of 
32 or 64 optical or tape backup devices. All the drive elements within an HSM system are 
confi gured as a single drive array (a bit like RAID). 
Details about various tape rotations are beyond the scope of this book, 
but if you want to learn more about them, search by their names on the 
internet.