Distributed Databases

Distributed Databases

Data does not have to be stored in the same location as it is used. In an organisation with many different offices, each office will need access to data. Where many locations require access to the same database then a distributed database is used.

Duplication

• This is where each location that requires data has an entire copy of the database in its own location

Central database with local indexes

• An index is data structure that works like the index in a book. It contains references to the data that allow it to be looked up and accessed quickly

• The database is held in a central location and the indexes, either those specific to a particular office/shop or all of them, are held locally. This allows each individual lovsation to perform processing locally and to use the central database for retrieval of a specific record.

Partitioning

• This is where parts of the data are stored at different locations. No single locstiong stores all the data

• There are two types of partitioning: vertical and horizontal

Vertical Partitioning

• This is where the data is partitioned based on the individual fields. You end up with a new table that contains only those fields that are not required.

• An office for example, may only be given access to certain fields within the date. For example, pay fields may not be accessible. Normalisation is a type of vertical partitioning

Horizontal Partitioning

This is where the date is partitioned based pn records. The data is split into different tables and each table contains exactly the same structure. However, each table will contain different rows. This can be used where only certain people/offices need data that is relevant to them.

Use of distributed databases

Comparison

• Speed - How quick is it go run queries and obtain the date that is required

• Security - Keeping the data secure, backed up and maintaining the integrity and consistency of the data.

• Flexibility - changing the date structure and queries to meet the needs of an organisation

• Storage space required and communication links

Duplicated

Speed is high because of local access

Security is down to the individual locations - multiple copies of tgevdats govr access points. Backing up is the responsibility of individuals but there are multiple copies if something goes wrong

Data integrity and consistency are major problem as a change in one database needs to be reflected throughout all of them. There are also more potential points of access for viruses

Every location needs a complete copy of the database - this increases the amount of storage and backup space required. Communication links are required to perform updates to each site. If the link is down, the integrity and consistency of the data suffers

Local Index

Where a local index exists speed is relatively high. Only data is transferred, no processing is required centrally. If the query and report is not available locally, then the speed is slower. Security is…