A data recovery technique for Redis using internal dictionary structure

Abstract

Since database systems are vulnerable to attacks the research in database forensics is gaining more importance. Database systems are used to store sensitive and confidential information. The loss of vital information from database can prove to be costly for any organization which may happen due to accidental or purposeful data deletion. Data recovery research in NoSQL databases is extremely important due to rapid growth of big data applications. The key-value database Redis, is one of the popular NoSQL database systems used by several applications. This article presents a data recovery technique for Redis database using internal dictionary structure. Redis is in-memory database and hence it does not leave any kind of signature of data deletion on the disk except log file. The proposed technique is based on tracing the memory addresses of dictionary data structure where key-value pairs are stored. The performance of this technique is evaluated by conducting several experiments on a sample dataset. The technique is found to be effective for the recovery of string, list, hash, set and sorted set data types.