Advantages and disadvantages of distributed systems?

Answer 1

Advantage:

Improved scalability

Unlike a single-database system, in which the amount of data that can be stored depends on the limitations of one host and one database, a distributed-database system is easily scalable and, therefore, set up for growth. As more sites or regions become part of the Teamcenter Enterprise network, you can expand the database topology to include new databases.

With multiple databases, Teamcenter Enterprise data is divided into logical pieces, so that users usually work only with the data that is most applicable to them.

Improved performance

The majority of database inserts, queries, updates, and deletions are on user data; therefore, a distributed-database environment separates user data from centralized or shared data and stores it locally. Local user databases reduce network traffic and eliminate network bottlenecks on most transactions.

Local user databases also distribute the user load in terms of system resources, such as memory, I/O, and disk. In addition, a distributed-database environment separates the user data from the data related to operational or background processing, which also reduces the amount of local system resources used.

Increased availability

Because user databases are independent, if one database is unavailable, users of other user databases can continue to work.

In a distributed-database environment, selected classes are replicated in other databases when they are created, deleted, or updated. This replication increases availability. For example, by replicating administrative data to a local host, chances are increased that the data can be read when it is needed.

By separating data, administrators have more flexibility in determining the frequency and types of backups needed for different types of data. This also increases availability.

Disadvantage:

Degradation of performance on a small network

One user action can cause activity in several databases, some of which may be remote. The additional overhead of these transactions can be a performance penalty when the total amount of data in the network is small.

Users also see slower performance when accessing user data that is not local. For this reason, the distribution of user data must be carefully planned to balance the need for local data with the desire for fast access to a large amount of data.

Increased use of database space

The schema of all databases must be the same, that is, every table must exist in every database. Therefore, database space is used for tables that may never be accessed. When the number of tables is very large, the amount of space used this way can be significant. Administrators must use database storage parameters to size tables and reduce database space consumption.

Complex use and administration

The use and administration of distributed databases are more complex than for a single-database network:

• Users must be aware of the concept of database scope and where different types of data are stored.

• Administrators must keep the schemas of all databases synchronized and ensure that the network is configured to take optimal advantage of the distributed-database environment in terms of availability and performance.

  Although database backups are more flexible in a distributed-database environment, multiple databases add complexity to the backup process.