MySQL Atchitecture and Concepts

MySQL Architectures And Concept(s)
Tuyen Vuong – Web Developer
vdt[dot]hutech[at]gmail[dot]com
MySQL Research
01-Aug-2013

MySQL Architecture | 01-Aug-2013 | 2
MySQL Architecture Overview

APPLICATION LAYER
 Users and clients interacts with the MySQL RDBMS.
 There are three components in this layer.
• Administrators
• Clients
• Query Users
 Query users interact with MySQL RDBMS using “mysql”.
 Administrators use various administrative interface and utilities like
mysqladmin, isamchk etc.
 Clients communicate with the MySQL RDBMS through various
interfaces and utilities like the MySQL APIs.
 “mysql”is actually a query interface. It’s a monitor that allows users
to issue SQL statements and view the results.

LOGICAL LAYER
 The logical layer of MySQL architecture is divided
into various subsystems.
• Query Processor.
• Transaction Management.
• Recovery Management.
• Storage Management.
 The above mentioned sub systems work together to
process the requests issued to the MySQL
database server.

MySQL’s Logical Architecture

Query Processor
• Query Processor further consists of the following systems.
 Embedded DML Precompiler.
 DDL Compiler.
 Query Parser.
 Query Preprocessor.
 Security/Integration Manager.
 Query Optimizer.
 Execution Engine.
• The output of one of the above component becomes the input for
another.
• Query processor layer is scalable and evolvable.

Transaction Management
 It facilitates concurrent data access.
 Provides locking facility.
 Ensures multiple users/sessions access data
simultaneously in a consistent way.
 Prevents data corruption or data damage.
 Lock Manager is the sub component name that
handles locking.

Recovery Management
• Log Manager
 Logs every operation executed in the database.
 Stores the operations logs as MySQL Commands.
 Incase of SYSTEM crash executing these
commands will bring back the database to its last
stable state.
• Recovery Manager
 Responsible for recovering the database to its last
stable state.
 Uses the logs created by the log manager.

Storage Management
• Storage Manager
 It acts like an interface with the OS.
 Its main job is efficient data write to the disk.
 Storage Manager writes to disk all of the data in the user tables,
indexes, logs as well as the internal system data.
• Buffer Manager
 It allocated memory resources.
 It decides
• Resource Manager
 Accepts the requests from execution engine.
 Requests the details from buffer manager.
 It actually receives references of data with memory from buffer
manager.
 Returns this data to the upper layer.

MySQL’s - Concept
Concurrency Control

MySQL’s - Concept
Lock Types and Lock Level
• Lock Types:
 Read Lock - The locked data is reserved for read by the current session.
Other sessions can read the locked data. But they can not write (update)
the locked data. A read lock is also called a shared lock.
 Write Lock - The locked data is reserved for write by the current
session. Other sessions can not read and write the locked data. A write
lock is also called an exclusive lock.
• Lock Level:
 Table Lock - The lock is set at the table level. All rows in the locked table
are locked.
 Row Lock - The lock is set at the row level. Some rows of a table are
locked. But other rows are not locked.
 Column Lock - The lock is set at the column level. Some columns of a
row are locked. But other columns are not locked.

MySQL’s - Concept
Table Locks
 When a user holds a WRITE LOCK on a table, no
other users can read or write to that table
 When a user holds a READ LOCK on a table, other
users can also read or hold a READ LOCK, but no
user can write or hold a WRITE LOCK on that
table.
 For example, if a user holds a WRITE LOCK on a
table, no other user can issue
a SELECT, UPDATE, INSERT, DELETE, or LOCK
operation on that table.

MySQL’s - Concept
Row Locks
Why this needed?
 Main reason to use these locking is to handle the concurrent
requests in proper way. This is the must required features when you
are dealing with important data just like financial details.
Pre Check Before Read Locking
 First of your table’s storage engine must be set as InnoDB.
 Your row locking query must executes after starting the transaction.
Conclusion
 It is better to use row locking mechanism if your database has high
volume of insert and update statements. But few thing to keep in
mind is that your table storage is set as InndoDB and your query
must executes after starting the transaction.

MySQL’s - Concept
Transactions

MySQL’s - Concept
Transactions
• A transaction is a sequential group of database
manipulation operations, which is performed as if it
were one single work unit. In other words, a
transaction will never be complete unless each
individual operation within the group is successful. If
any operation within the transaction fails, the entire
transaction will fail.
• Practically you will club many SQL queries into a
group and you will execute all of them together as a
part of a transaction.

MySQL’s - Concept
Properties of Transaction
•Atomicity: ensures that all operations within the work unit
are completed successfully; otherwise, the transaction is
aborted at the point of failure, and previous operations are
rolled back to their former state.
•Consistency: ensures that the database properly changes
states upon a successfully committed transaction.
•Isolation: enables transactions to operate independently of
and transparent to each other.
•Durability: ensures that the result or effect of a committed
transaction persists in case of a system failure.

MySQL’s - Concept
Transaction – COMMIT and ROLLBACK
•When a successful transaction is completed, the COMMIT
command should be issued so that the changes to all involved
tables will take effect.
•If a failure occurs, a ROLLBACK command should be issued to
return every table referenced in the transaction to its previous state.
•If AUTOCOMMIT is set to 1 (the default), then each SQL statement
(within a transaction or not) is considered a complete transaction,
and committed by default when it finishes. When AUTOCOMMIT is
set to 0, by issuing the SET AUTOCOMMIT=0 command, the
subsequent series of statements acts like a transaction, and no
activities are committed until an explicit COMMIT statement is
issued.

MySQL’s - Concept
Isolation Levels

MySQL’s - Concept
Deadlocks

MySQL’s - Concept
Multiversion Concurrency Control
•MVCC (Multi Version Concurrency Control) achieves
both of concurrent update and isolation.
•While execute update transaction, MVCC doesn’t
overwrite but create new version.
•First query (including SELECT query) in transaction
uses newest version. Continued SELECT queries sees
the version same to first query.

MySQL’s - Concept
MySQL’s Storage Engines

MySQL’s - Concept
InnoDB
• InnoDB tables are transaction-safe (ACID compliant) storage
engine that has commit, rollback, and crash recovery capabilities for
data protection.
• It supports row-level locking.
• Foreign key referential-integrity constraint can be defined.
• Table can extend to any size even beyond 2 GB and power loss
recovery is fast.
• The InnoDB stores user data in clustered indexes
• This reduces I/O for common queries based on primary keys
• InnoDB should be used for applications requiring the data
integrity.

MySQL’s - Concept
MyISAM (1)
•MyISAM is the improved version of the original storage engine
of MySQL, ISAM
•After MySQL 3.23, MyISAM replaced ISAM as the default
storage engine.
•The MyISAM engine is fast and thus, preferred for web and
other application environments.
•It is also used for data warehousing
•MyISAM is not transaction-safe and supports 64 keys per
table with maximum key length of 1024 bytes
•The size of MyISAM table depends on the host operating
system

MySQL’s - Concept
MyISAM (2)
•MyISAM table allows table level locking only.
•There are no limitations on data file transfer and the data files
can be ported from system to system
•The foreign key constraint cannot be defined
•MyISAM is the only storage engine that supports Full-text
search
•It also supports one auto increment column per table
•A high-byte-first pattern for saving numeric key values ensures
faster indexing
•It can be used where fulltext indexing is needed

MySQL’s - Architecture
Reference:
 Oreilly.High.Performance.MySQL.3rd.Edition
 MySQL Conceptual Architecture
 http://www.mysql.com
 Internet Bloggers

Questions?
vdt.hutech@gmail.com
website: tuyenvuong.info

MySQL Atchitecture and Concepts

Recommended

More Related Content

What's hot (20)

Similar to MySQL Atchitecture and Concepts (20)

Recently uploaded (20)

MySQL Atchitecture and Concepts