Security of distributed databases

Security of distributed databases

1. Interception of data

   • gathering of data by a third party (e.g. hackers) whilst it sent from one location to another
   • all distributed databases move data between themselves and this gives an opportunity for interception.
   • some distributed databases have more access points which enables them to intercept the data easier
   • common way of gathering data is though the communication links which can be wired or wireless
   • interception usually occurs during the communication of data. partitioned and centralised databases makes greater use of communication links and so could be at a greater risk of having data intercepted
   • duplicated database uses communication links when it is being updated.however might be frequently updated. so increase of communication links leads to high risk of interception
   • Encryption doesnt prevent the data from being intercepted but prevents the person from understanding the intercepted data

2. Physical access of data

   • preventing physical access to data is about preventing the individual from getting to the terminal
   • partitioned and replicated distributed databases stored in many locations - so many location where data can be accessed
   • Physical security methods;
     • Security guards
     • camera
     • card scanner
     • access control on entrances (locks on doors)
     • eye scanner
   • Software security methods;
     • encryption
     • firewall
     • backup
     • log in procedures - username and passwork
     • scan
     • update
     • user access levels
   • so those location needs a physical security and maintain the high level of security

3. Consistency and integrity of data

   • Consistency of data
     • in partitioned databases - consistency is a major problem; changes held in one location needs to be reflected in all of the databases from different locations
     • makes sure that the copied data that is stored in different location is the same
     • in vertical - this can be done through primary and foreign keys. the consistency of data will also be enforced through referential integrity, cascade and update deletions
     • in horizontal - consistency is easy to maintain if each piece of data (the records) is stored at a single location. however. if the same data is duplicated then its difficult
       • a record-locker system could be implemented to maintain the consistency across the different locations
     • its easier to maintain in a centralised databases with local index. all the data is stored at one location (centrally) so any changes can be completed at the central location
       • updates should be completed on a regular basis and might be completed each time to the central database
   • Integrity of data
     • makes sure that the data stored is complete and whole
     • incomplete data can cause issues when the data needs to be retrieved
     • the data entered into distributed database system must be valid and varified
     • Validation
       • automatic checking of the data entered into the distributed database system
       • ensures that the data entered conforms to rules and sensible and reasonable
       • validation checks
         • length check - ensure that the data is reasonable length by using preset boundaries
         • integrity - confirms the value of a piece of data by comparing with other stored data
         • Range - ensures the data within a preset range (lower and upper case is set and it should be in the boundries)
     • Varification
       • ensures that the data entered is the same as the source
       • can also be used to confirm the integrity of the data as its copied between different parts of the distributed database
       • data should not be changed while copying or replication as it can cause errors
       • verification checks
         • double entry - ensures that the data entered is accurate. data is entered twice and compared by the system e.g. typing email address twice
         • screen verification - when the data has been entered, it displayed on-screen: the user checks the data and performs an action to confirm it is correct
   • difficult to main consistency and integrity of data in a duplicated distributed databases; each location will make changes to the copy of the database it hold, but general synchronisation is carried out.