Scribd
Upload
2 views3 pages

Data Replication

The document discusses data replication in distributed databases, outlining three cases: fully replicated, no replication, and partial replication. Fully replicated systems improve availability and performance but slow down updates, while no replication simplifies design but suffers from poor fault tolerance. Partial replication offers a balance of performance and cost but requires careful management to avoid inconsistencies in availability.

Uploaded by

samali13x
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
2 views3 pages

Data Replication

The document discusses data replication in distributed databases, outlining three cases: fully replicated, no replication, and partial replication. Fully replicated systems improve availability and performance but slow down updates, while no replication simplifies design but suffers from poor fault tolerance. Partial replication offers a balance of performance and cost but requires careful management to avoid inconsistencies in availability.

Uploaded by

samali13x
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Data Replication:

• Replication If a fragment is stored at more than one site


• Replication is useful in improving the availability of data.
Three cases for Replication:
1. Fully replicated distributed database
2. No replication
3. Partial replication
1- Fully replicated distributed database :
Replication of the whole database at every site in the distributed
system
Advantages:
• improve availability remarkably
• improves performance for global queries
Disadvantages:
• slow down update operations
• makes the concurrency control and recovery techniques more
expensive
Examples:
• Systems with high availability requirements where
transactions are primarily .
• Banking systems where all branches keep full copies of
customer data to ensure constant service.
2- No replication of the data :
Each fragment is stored at exactly one site
Advantages:
• Simple System Design: The architecture is easier to design as
there are no multiple copies to manage.
• Faster Updates: Updates are quick because only one location
needs to be modified.
Disadvantages:
• Poor Fault Tolerance: The system is highly vulnerable to
node failures.
• Slower Queries: Performance drops if a site frequently needs
to access data stored at a remote site.
Examples:
• Systems where updates are very frequent, making the cost of
replication too high.
3- Partial replication of the data:
• The number of copies of each fragment can range from one up
to the total number of sites in the distributed system
Advantages:
• Balance: It provides a balance between performance, cost, and
availability.
• Reduced Update Overhead: It is more efficient for updates
compared to a fully replicated system.
Disadvantages:
• Complex Design: It requires careful decision-making and
management to determine which fragments to replicate and
where to place them.
• Inconsistent Availability: Some data may still be unavailable if
it is not among the replicated fragments.
Examples:
• E-commerce systems that replicate product information for
fast browsing but keep transaction logs in a single location.

You might also like