CH FOUR

1. Basic concept of a Transaction,

A database transaction is a sequence of database operations treated as a single unit, ensuring all-or-nothing
execution, data integrity, and reliability.

2. Transaction Management
Transaction management in databases ensures that database operations are executed reliably and
consistently, following the ACID properties: Atomicity, Consistency, Isolation, and Durability.

3. Properties of Transactions,
The properties of transactions, often abbreviated as ACID, are:
Atomicity: Transactions are all-or-nothing; either all operations within a transaction are completed
successfully, or none of them are. There are no partial executions.
Consistency: Transactions ensure that the database remains in a consistent state before and after the
transaction. Integrity constraints and data validity rules are maintained.
Isolation: Transactions are isolated from each other to prevent interference. Changes made by one
transaction are not visible to other transactions until they are committed.
Durability: Once a transaction is committed, its changes are permanent and survive system failures. They are
stored in a way that ensures they can be recovered even after a crash.

4. Concept of Schedule
A schedule in a database refers to the chronological order of transactions' execution. It shows how
transactions are interleaved and executed concurrently in a multi-user database system. The schedule
reflects the sequence of read and write operations performed by transactions, influencing data consistency
and concurrency control.

5. Serial Schedule

A serial schedule in a database executes transactions sequentially, one after another, without interleaving
or concurrency.

6. Serializability: Conflict and View

Serializability: Guarantees equivalent outcome to serial execution.
confict Serializability: Conflict serializability ensures transactions can be ordered to prevent conflicts (e.g.,
read-write, write-write) for consistency.
View Serializability: View serializability ensures that different transaction schedules produce the same final
database state, preserving consistency regardless of the order of execution.

7. Cascaded Aborts

Cascaded aborts occur when a transaction aborts, triggering the rollback of other dependent
transactions, potentially leading to a chain reaction of aborts throughout the system.

8. Recoverable and Nonrecoverable Schedules

Recoverable Schedules: Ensure that if a transaction commits, all transactions it depends on have also
committed. This guarantees that no transaction reads data written by another transaction that later aborts.
Nonrecoverable Schedules: Allow transactions to read data from transactions that haven't committed yet. If a
transaction later aborts, the changes it made may have been read by other transactions, leading to
inconsistencies.
9. Concurrency Control: Need, Locking Methods,

Concurrency Control:Concurrency control manages simultaneous access to shared data in a multi-user

database system, ensuring consistency and preventing interference between transactions executing
concurrently.

Need : Concurrency control is essential in multi-user database systems to manage simultaneous access to
shared data, preventing interference and maintaining consistency.

10. Deadlocks
Deadlocks in a database occur when two or more transactions are waiting for each other to release resources
that they need, resulting in a situation where no transaction can proceed.

11. Timestamping Methods

Timestamping methods in databases assign a unique timestamp to each transaction, typically based on its
start time or order of arrival. These timestamps are then used to determine the serializability of transactions
and resolve conflicts in concurrent execution.

12. Recovery methods : Shadow-Paging and Log-Based Recovery

Shadow-Paging: Maintains a shadow copy for recovery by updating a separate set of pages before
committing.
Log-Based Recovery: Uses a transaction log to record changes, enabling undo or redo operations for recovery
after a crash.

13. Checkpoints
Checkpoints record consistent database states for faster recovery after crashes.

14. Query Processing

Query processing: Converts user queries to executable instructions for efficient database data retrieval.

15. Query Optimization

Query optimization: Selects the most efficient execution plan for processing queries, minimizing resource
usage and execution time.

16. Performance Tuning.

Performance tuning optimizes system performance by adjusting various parameters, configurations, and
algorithms to enhance efficiency and throughput, ensuring optimal use of resources.
CH FIVE
1. Parallel and Distributed Databases
Parallel databases distribute data processing tasks across multiple processors to improve performance, while
distributed databases store data across multiple nodes connected via a network for scalability and fault
tolerance.

2. Introduction to Database Architectures

Database architectures refer to the design and structure of database systems, including how data is
organized, stored, accessed, and managed, to meet specific requirements such as performance, scalability,
and reliability.

3. Multi-user DBMS Architectures,

Multi-user DBMS architectures support concurrent access to a database by multiple users, typically utilizing
client-server or peer-to-peer models, ensuring data consistency, security, and efficient resource utilization.

4. Parallel Databases:
Parallel databases distribute data processing tasks across multiple processors to improve performance by
leveraging parallelism and exploiting parallel hardware architectures.

5. Speedup and Scale up,

Speedup refers to the improvement in performance achieved by parallelizing tasks across multiple processors
or nodes, while scale-up involves increasing the capacity or capability of a single system or component to
handle larger workloads.

6. Architectures of Parallel Databases.

Parallel database architectures distribute data and processing tasks across multiple nodes or processors to
enhance performance, typically employing shared-nothing or shared-disk architectures for efficient
parallelization.

7. Distributed Databases
Distributed databases store data across multiple nodes connected via a network, enabling scalability, fault
tolerance, and decentralized data management.

8. Architecture of Distributed Databases

Distributed database architectures distribute data across multiple nodes connected via a network, often
utilizing replication, partitioning, or distributed transaction management for scalability and fault tolerance.

9. Distributed Database Design

Distributed database design involves partitioning, replication, and allocation strategies to distribute data
across nodes in a network, ensuring efficiency, fault tolerance, and scalability.

10. Distributed Transaction:

Distributed transactions involve multiple operations across distributed databases, ensuring consistency,
atomicity, isolation, and durability (ACID properties) despite network and system failures.

11. Basics
Distributed transactions are fundamental units of work that span multiple databases or systems, maintaining
ACID properties (Atomicity, Consistency, Isolation, Durability) across distributed environments.
12. Failure modes,
Failure modes in distributed transactions refer to potential issues such as network failures, system crashes, or
communication errors that can disrupt the coordination and execution of transactions across multiple
databases or systems.

13. Commit Protocols,

Commit protocols in distributed transactions define rules and procedures for coordinating the commit or
rollback of transactions across multiple distributed nodes to ensure data consistency and integrity.

14. Concurrency Control in Distributed Database.

Concurrency control in distributed databases manages simultaneous access to shared data across multiple
nodes, ensuring consistency, isolation, and integrity despite concurrent transactions executing on different
nodes.
CH SIX
1. NoSQL Database
NoSQL databases are non-relational databases optimized for handling large volumes of unstructured or semi-
structured data, offering flexibility, scalability, and high performance for diverse use cases.

2. Types and examples of NoSQL Database- Key value store,

Key-value stores are NoSQL databases that store data as a collection of key-value pairs, where each unique
key is associated with a value. Examples include Redis, Amazon DynamoDB, and Apache Cassandra.

3. document store,
Document stores are NoSQL databases that store data in a semi-structured format, typically using JSON or
BSON documents. Each document is a self-contained unit containing key-value pairs, allowing for flexible
schema and nested data structures. Examples include MongoDB, Couchbase, and Elasticsearch.

4. graph,
Graph databases are NoSQL databases that represent data as nodes, edges, and properties, enabling efficient
storage and traversal of complex relationships between entities. Examples include Neo4j, Amazon Neptune,
and JanusGraph.

5. Performance
Performance in NoSQL databases is characterized by their ability to handle large volumes of data, high
throughput, low latency, and scalability.

6. Structured verses unstructured data

Structured data follows a fixed schema, while unstructured data lacks a predefined structure. NoSQL
databases are adept at handling unstructured data due to their flexible schema and scalability.

7. Distributed Database Model

The distributed database model involves distributing data across multiple nodes in a network, providing
scalability, fault tolerance, and decentralized data management.

8. CAP theorem and BASE Properties,

CAP theorem: Consistency, Availability, Partition tolerance trade-off in distributed systems.
BASE properties: Basically Available, Soft state, Eventually consistent approach in distributed databases.

9. Comparative study of SQL and NoSQL,

SQL databases are relational, structured, and suitable for complex transactions and queries. NoSQL
databases are non-relational, flexible, and efficient for handling unstructured or rapidly changing data. The
choice depends on specific use cases, scalability needs, and data requirements.

10. NoSQL Data Models: NoSQL data models include key-value, document-oriented, column-family, and graph
databases, each tailored to specific data storage and retrieval needs.

11. Case Study-unstructured data from social media: Analyzing unstructured data from social media platforms
involves collecting, storing, and analyzing text, images, videos, and other content to extract insights about
user behavior, sentiment analysis, and trends.

12. Introduction to Big Data: Big Data refers to large volumes of structured, semi-structured, and unstructured
data that cannot be processed or analyzed using traditional database management tools. It encompasses the
three Vs: volume, velocity, and variety.
13. HADOOP: Hadoop is an open-source framework for distributed storage and processing of Big Data across
clusters of commodity hardware. It provides scalability, fault tolerance, and parallel processing capabilities
for handling massive datasets.

14. HDFS (Hadoop Distributed File System): HDFS is the primary storage system used by Hadoop for distributed
storage of large datasets across multiple nodes in a Hadoop cluster. It stores data in a fault-tolerant manner
and enables high-throughput data access.

15. MapReduce: MapReduce is a programming model and processing engine used in Hadoop for parallel
processing and analysis of large datasets. It divides tasks into map and reduce phases, enabling distributed
computation across a Hadoop cluster.