Types of locks, Database Management System

Types of Locks

There are two necessary types of locks:

  • Binary lock: This locking mechanism has two states for to a data item: locked or unlocked
  • Multiple-mode locks: In this locking type every data item can be in 3 states read locked or shared locked, write locked or exclusive locked or unlocked.

Let us first take an instance for binary locking and describes how it solves the concurrency related troubles. Let us reconsider the transactions T1 and T2 for this purpose; though we will add to needed binary locks to them.

Schedule

T1

T2

Lock X

Lock X

 

Read X

Read X

 

Subtract 100

Subtract 100

 

Write X

Write X

 

Unlock X

Unlock X

 

Lock X

 

Lock X

Lock Y

 

Lock Y

Read X

 

Read X

Read Y

 

Read Y

Display X+Y

 

Display X+Y

Unlock X

 

Unlock X

Unlock Y

 

Unlock Y

Lock Y

Lock Y

 

Read Y

Read Y

 

Add 100

Add 100

 

Write Y

Write Y

 

Unlock Y

Unlock Y

 

 

 

 

                                                         Figure: An incorrect locking implementation

Does the locking as completed above solve the problem of concurrent transactions? No the similar problems still remains. Try working with the old value. Therefore, locking should be completed with some logic in order to make sure that locking results in no concurrency related problem. One such answer is given below:

Schedule

T1

T2

Lock X

Lock X

 

Lock Y

Lock Y

 

Read X

Read X

 

Subtract 100

Subtract 100

 

Write X

Write X

 

Lock X (issued by

T2)

Lock X: denied as T1 holds the lock.

The transaction T2 Waits and T1 continues.

Read Y

Read Y

 

Add 100

Add 100

 

Write Y

Write Y

 

Unlock X

Unlock X

 

 

The lock request of T2 on X can now

be granted it can resumes by locking X

Unlock Y

Unlock Y

 

Lock Y

 

Lock Y

Read X

 

Read X

Read Y

 

Read Y

Display X+Y

 

Display X+Y

Unlock X

 

Unlock X

Unlock Y

 

Unlock Y

                                                              Figure: A correct but restrictive locking implementation

Therefore, the locking as above when you get all the locks at the starting of the transaction and release them at the end ensures that transactions are implemented with no concurrency related problems. Though, such a scheme limits the concurrency. We will talk about a two-phase locking method in the next subsection that gives sufficient concurrency. Though, let us first discuss multiple mode locks.

Posted Date: 3/12/2013 2:36:18 AM | Location : United States







Related Discussions:- Types of locks, Assignment Help, Ask Question on Types of locks, Get Answer, Expert's Help, Types of locks Discussions

Write discussion on Types of locks
Your posts are moderated
Related Questions
What is Data dictionary? A data dictionary is a reserved space in a database that is employed to store information about the database itself. A data dictionary may consist o

Why did SAP introduce the extended star schema? Explain why it is reported to be better than the traditional schema model?

Explain P command in the respect of QBE? P: It is the command in which is used to print (logically display) the value of the attribute in whose cell it is writ

1. Which model is used to store data in tree structures Ans: Hierarchical model is used to store data in tree structures

Illustrate the class diagram for class room scheduling system If in any college, the number of classrooms are limited which have to be allocated to various classes and instruct

What are the advantages of object ID? The main advantages of Object IDs are as below: i) IDs are not changing  ii) IDs are completely independent of the changes in data v

Single Valued Normalisation  Codd in the year 1972 formed three normal forms (1NF, 2NF, and 3NF). These were based on functional dependencies between the attributes of a relati

Define Serializable schedule Serializable schedule: An interleaved schedule of much more than one transaction is known as a serializable schedule, if it is equal to some serial

Consider the following frequency counts of some itemsets in a transaction database r : freq({A}, r) = 0.405 freq({B}, r) = 0.510 freq({C}, r) = 0.303 freq({A

What are the three data anomalies that are similarly to take place as a result of data redundancy?  Can data redundancy be completely eliminated in database approach?  Why or why n