Disaster recovery techniques for database systems

Recovery techniques are an important factor in the database management systems. The database recovery system is a highly complex system composed of many transactions that are being executed each second. The system may undergo different types of failure including;


  1. Transaction failure

When a transaction fails to execute, it has to abort. The failure maybe as a result of:

  • Logical errors

This refers to any internal error such as code errors that make a transaction to be incomplete.


This is where the database management system aborts an active transaction due to a system failure condition such as deadlock.


  1. Disk failure

Disk failure attributed to, when the hard-disk drive fails due to disk head crash and bad sectors among others. This annihilates the disk partially or fully.


  1. System crash

External errors such as power supply shortage may lead to a sudden stop in the system activities which may cause the system to crash thus leading to loss of certain data that had already been executed.


When the system undergoes such failures, it has to undergo some sort of techniques to recover the lost data. Such techniques may include:


  1. Incremental backup

This is an efficient and reliable way of backing up your data. Files that have been changed are copied to archive after a transaction is completed or at intervals. In incremental backup, not all files are backed up, it’s only the once that have been modified overtime.


Incremental backup is fast and reduces the storage space since the only backed up data it consists is the modified files. During recovery of data, it provides grounds whereby data can be recovered to a more precise point in time. However, it takes long to restore data

  1. Audit trail

This refers to a series of activities on log files that are recorded. This technique provides security since it records all activities from all users. If data is lost, the audit trail, gives track of track accessed files, whether complete or attempted. It is essential in recovering data since it traces the transactions that have altered the log records.


  1. Differential files

Differential files refers to the differential backup type in which data is preserved and only the difference in data is saved since the last full back up. Differential back is quicker in terms of data recovery time since it only requires a full back up and the last differential backup to restore the lost data.


  1. Salvation program

This technique is applied when all other methods of data recovery fail. It is often put into effect after a system crash so as to restore the system to a valid state and is normally applicable where in buffers were lost in a system crash. Salvation Data has data recovery tools that prevent data loss and also necessitate high success rate in data recovery.


  1. Multiple copies

During the database transaction operations, multiple copies of each active files are maintained. The essence of this is to ensure that, in case of failure, different versions of files are available and can be recovered easily by making comparisons between the versions that can help in finding a consistent state version of the file. In this case, a valid state version of a file can be restored easily.

These techniques are very essential in data restoration in case of a disaster. However, not all data may be recovered but at least some of it is guaranteed.


I'm Elizabeth part time blogger and working as Software Testing Head.