DBMS ASSIGNMENT 2 Q3

Submitted By: Afra Rasheed (SE-21007)

Explore the different levels of RAID, understand and depict them.

The hardware running the DBMS should be designed to handle faults effectively, ensuring the system
continues to operate even if individual hardware components fail. This involves incorporating redundant
components that seamlessly integrate into the system when failures occur. The key hardware components
requiring fault tolerance include disk drives, disk controllers, CPU, power supplies, and cooling fans.
Among these components, disk drives are particularly vulnerable with shorter time intervals between
failures compared to others.
One approach to address these challenges is using Redundant Array of Independent Disks (RAID)
technology. RAID involves creating a large disk array consisting of multiple independent disks organized
to enhance reliability and performance. Data striping is employed, dividing data into equal-sized
partitions and transparently distributing them across multiple disks. This creates the illusion of a single,
fast disk while actually distributing the data across smaller disks. Striping improves overall I/O
performance by allowing parallel servicing of multiple I/Os and load balancing across disks.
Reliability in RAID is enhanced by storing redundant information using parity or error-correcting
schemes. Parity schemes use parity bits associated with each byte to detect corruption, while error-
correcting schemes use additional bits to reconstruct data if a single bit becomes corrupt. Different RAID
levels exist, each offering different trade-offs between performance and data protection.

Some common RAID levels include:

- RAID 0: Non-redundant, offering improved performance through data striping.
- RAID 1: Mirrored, maintaining identical copies of data on different disks for increased redundancy.
- RAID 0+1: Combines striping and mirroring.
- RAID 2: Uses memory-style error-correcting codes with a single-bit striping unit.
- RAID 3: Utilizes bit-interleaved parity for redundancy.
- RAID 4: Implements block-interleaved parity with separate parity disks.
- RAID 5: Uses block-interleaved distributed parity, striping parity data across all disks.
- RAID 6: Provides additional redundant data for protection against multiple disk failures.

The choice of RAID level depends on factors such as performance requirements and data protection
needs.
1. Web servers: For web servers that require high performance and data availability, RAID 10 (RAID
1+0) is commonly recommended. RAID 10 combines both mirroring (RAID 1) and striping (RAID 0)
techniques to offer excellent performance and redundancy. It provides both improved read and write
speeds along with data protection.
2. Video editing and multimedia production: In environments where large video files and multimedia
content are frequently accessed and edited, RAID 5 or RAID 6 can be suitable choices. These RAID
levels provide a good balance between storage capacity, performance, and data protection. RAID 5 is
commonly used when there are three or more disks, while RAID 6 offers enhanced protection against
multiple disk failures.
3. Database servers: Database servers often require a combination of performance and data protection.
RAID 10 is a popular choice for critical database systems as it offers excellent read and write
performance with redundancy. RAID 10 provides high availability and fast recovery in case of disk
failures.
4. Archival and backup storage: For long-term data storage and backup purposes, RAID 6 is often
recommended. RAID 6 provides a high level of data protection and fault tolerance by allowing for the
failure of two disks simultaneously. This ensures that data remains intact even during a disk failure and
helps minimize the risk of data loss.

Advantages of RAID:
1. Fault tolerance: RAID provides fault tolerance by distributing data across multiple disks and
incorporating redundancy. This means that if one disk fails, the data can still be accessed from other disks,
ensuring data availability and minimizing the risk of data loss.
2. Improved performance: RAID can enhance overall system performance through techniques like data
striping, which allows data to be split across multiple disks and accessed in parallel. This parallel access
can significantly improve read and write speeds, especially for large files or in high-demand
environments.
3. Increased storage capacity: RAID configurations can combine the storage capacity of multiple disks,
providing a larger overall storage space. This can be beneficial when dealing with large amounts of data
that require ample storage capacity.
4. Data protection: With the use of redundancy mechanisms, such as parity or error-correcting schemes,
RAID can protect against data corruption or loss. These schemes enable the reconstruction of lost or
corrupted data using the redundant information stored across the disks.

Disadvantages of RAID:
1. Cost: Implementing RAID can involve additional hardware costs, including multiple disks, disk
controllers, and RAID controller cards. The cost can vary depending on the RAID level chosen and the
required storage capacity. RAID also requires additional power consumption and cooling resources,
which may increase operational costs.
2. Complexity: Setting up and configuring RAID can be complex, especially for more advanced RAID
levels. It requires a good understanding of RAID concepts, hardware compatibility, and proper
configuration to achieve optimal performance and data protection.
3. Performance trade-offs: While RAID can improve overall performance, certain RAID configurations,
especially those with redundancy mechanisms, can introduce overhead and impact write performance.
The calculation and storage of redundant information may slow down write operations, particularly in
RAID levels with higher levels of redundancy.
4. Limited fault tolerance: Although RAID provides redundancy and fault tolerance, it does not guarantee
complete protection against data loss or system failures. In certain cases, multiple disk failures or other
catastrophic events can still lead to data loss. Regular monitoring, maintenance, and backup strategies are
still essential to mitigate such risks.