An Efficient Backup Technique for Database Systems Based on Threshold Sharing Sahel Alouneh 1 , Sa'ed Abed 2 , Bassam Jamil Mohd 2 , Mazen Kharbutli 3 1 German Jordanian University, 2 Hashemite University, 3 Jordan University of Science and Technology {sahel.alouneh@gju.edu.jo, sabed@hu.edu.jo, Bassam@hu.edu.jo, kharbutli@just.edu.jo} Corresponding and principle author: Sahel Alouneh, Ph.D Abstract—Database security, corruption, and loss can be disruptive, time-consuming and expensive to organization operation and business continuity. Therefore, data protection and availability is a high priority and a sensitive concern during the design and implementation of information systems infrastructure. This paper proposes a novel technique for designing and implementing a database recovery and security system based on the threshold secret sharing scheme. Furthermore, a network-based database protection technique is devised and presented. Analysis of the proposed technique shows that it is effective and comprehensive while not imposing significant delay in order to secure, distribute and recover the data. Index Terms— Database; Threshold sharing; Security; Recovery I. INTRODUCTION AND MOTIVATION Business continuity necessitates the existence of a reliable database system. A database error or failure, even if for a short time period, can cause financial and social confusion, and may lead to the loss of valuable customer confidence [1, 2]. A database system failure may be caused by various reasons such as human errors or hardware failures. To ensure the reliability of the database in a system, the database management and recovery policy must be prepared beforehand prior to the occurrence of such errors and failures. In addition, the security of database content is of great concern, and in some situations is considered the first priority. This paper proposes a novel and efficient technique for designing and implementing a database protection system that is based on the Threshold Secret Sharing Scheme (TSSS) [3]. One main motivation behind this work is the ability to use the natural characteristics of the threshold secret sharing scheme in order to provide security and recovery for database systems. On the other hand, the original TSSS requires extra overhead for coding the shares and therefore this work suggests modification to the original TSSS to help in reducing this overhead size, especially when using this technique to recover from data loss and protecting the integrity of the data. The proposed technique relies on dividing the database into shares and distributing them to an array of hard drives (Sub-storage databases) in such a way that would make it possible to recover the original database even if not all hard drives are able to provide their shares. That is, the shares would contain enough redundancy making it possible for the whole database to be recovered if enough (but not all) hard drives are able to provide their shares. In addition, the proposed technique is also able to protect the confidentiality and integrity of the database system because the shares are coded. The proposed technique is analyzed and shown to be effective while imposing insignificant delays and overhead. The rest of the paper is organized as follows. Section 2 discusses the database protection background and related work. After that, Section 3 presents the proposed technique and Section 4 evaluates and compares it with other existing techniques. Finally, Section 5 concludes the paper. II. DATABASE PROTCTION BACKGROUND AND RELATED WORK Data recovery is the process of preserving data from damage or destruction. It may be required due to failures or errors that turn up in the form of physical damage to the storage device or logical damage to the file system that prevents it from being mounted by the host operating system. A primary storage medium used for data storage is the hard disk. Hard disks have higher failure rates relative to other storage mediums necessitating the existence of a backup and recovery system. A simple method that can be used to protect from hard disk failures is to make backup copies of the files on other mediums such as redundant hard disks, magnetic tapes, removable disks, or magnetic optical disks [2]. Currently, a popular method used to improve the reliability and performance of database storage on hard disks is RAID arrays (Redundant Arrays of Inexpensive Disks) [4]. RAID technology was developed to address the limitations and drawbacks of conventional disk storage systems in terms of fault-tolerance and performance. It can offer an improved fault tolerance and higher throughput levels compared to a single hard drive or a group of independent hard drives. While RAID arrays were once considered complex and expensive storage solutions, today they are easy to use, relatively inexpensive, and have become essential for a broad range of applications. In a RAID solution, multiple drives are organized in a single array, which is viewed by the operating system as a single disk. There are several different RAID "levels" or 2980 JOURNAL OF COMPUTERS, VOL. 8, NO. 11, NOVEMBER 2013 © 2013 ACADEMY PUBLISHER doi:10.4304/jcp.8.11.2980-2989