Database Concurrency

Download as pdf or txt
Download as pdf or txt
You are on page 1of 39

ADVANCE

DATABASE
SYSTEMS
DEVELOPMENT
[CC6001]
WEEK - 06
Database Concurrency
There are three sides of ACID.
ENHANCED LONG TERM MEMORY
DECREASED SHORT TERM MEMORY
AND I FORGOT THE THIRD
- TIMOTHY LARY
Requirement

MOST OF THE
CASES, DATABASE IS
ACCESSED BY
MULTIPLE USERS AT A
TIME
Concurrency Control
ENSURES THAT
CONCURRENT
OPERATIONS ARE
CARRIED OUT
 CORRECTLY
 EFFICIENTLY
Example
 Begin Transaction T1
 Read Balance1
 Balance1 = Balance1 – 100
 If Balance <0
 Print insufficient fund
 Abort T1
 End
 Write Balance1
 Read Balance2
 Balance2 = Balance2 +100
 Write Balance2
 Commit T1
Points to Remember

Effect of ABORT is to ROLLBACK the


TRANSACTION and UNDO changes it has made on
the DATABASE

In this example, TRANSACTION was NOT written


to the DATABASE prior to ABORT hence, NO
UNDO is necessary
Yet Another Scenario

Transactions TA and TB performs a series of


processing (sub-tasks) TA1, TA2, TA3 and TB1,
TB2, TB3 respectively.

CPU is shared by TA and TB and other


transactions in the system
Yet Another Scenario

Single Multi
Transaction Transaction
Environment Environment
Interleaved Transactions
The operations of the two transactions TA and TB are
said to be interleaved to achieve concurrent execution.

The operations of the two transactions TA and TB are


said to be interleaved to achieve concurrent execution.
Various concurrency problems may occur if proper control is not enforced!!!
Concurrency Problems

Lost Update

Violation of Integrity
Constraints

Inconsistent Retrieval
Lost Update
two concurrent transactions, say TA and TB, are
allowed to update an uncommitted change on the
same data item, say x.

The earlier update of x by Transaction TA is


overwritten by a subsequent update of x by
transaction TB.

Consequently, the update value of x by TA is lost


after the update by TB.
Lost Update - Example
Lost Update - Question

What should be the correct


balance after TA and TB?
Lost Update - Answer

400
Violation of Integrity Constraints
Integrity Constraint - Question

What should be the correct


balance after TA and TB?
Integrity Constraint - Answer

200
Inconsistent Retrieval

TA is summing up the total number of vacant teaching rooms in 3


building (Tower, Moorgate and Eden), whilst TB is moving 2
classes (2 rooms) from Tower to Eden.

Initial numbers of vacant rooms:


Tower: 10
Moorgate: 15
Eden 5
Total number 30
Inconsistent Retrieval
Inconsistent Retrieval
 Initial numbers of vacant rooms:
 Tower: 10
 Moorgate: 15
 Eden 5
 Correct numbers of vacant rooms after moving should be:
 Tower: 8 (10 – 2)
 Moorgate: 15
 Eden 7 (5+2)
 But we seem to have got 32 vacant rooms now!! WRONG
TOTAL!
Concurrency Control - Principles

Schedule multiple Maximize the


transactions in such a way degree of
as to avoid any concurrency or
interference between parallelism in the
them, while at the same system
time
Solutions to Concurrency Problems

Serialization
Constraints
must be
specified by
DBA using
Concurrency Control
appropriate
language

Transaction Scheduling
Serialization
ONE
Process Disadvantage
TRANSACTI
ON MUST
• Permits SERIAL • Inefficient for COMMIT
EXECUTION of MULTI USER
TRANSACTIONS environment BEFORE
ANOTHER
CAN START.
Concurrency Control

Process Mechanisms

• Allows CONCURRECT • Locking


EXECUTION of • Optimistic Scheduling
transactions in a • Time Stamping
CONTROLLED way.
Locking

Locking method is the most commonly


used approach to ensure serializability
of concurrent transactions.

Locking synchronizes interleaved


transactions so that it is
equivalent to some serial
execution of those transactions.
Locking - Idea

When a transaction requires the assurance that


some data will not change during the course of
processing (read/write) it, this transaction must
request and acquire a lock, in order to prevent
other transactions updating the data.
Lock Types
Exclusive Shared

• X (Write) • S (Read)
• Grant read/write access to a • Grant read only access to a
data item to the transaction data item to the transaction
which holds the lock which holds the lock
• Prevent any other transactions • Prevent any transaction
reading or writing the same writing data item.
data item. • Several transactions may hold
a S lock on the same data
item.
Locking Protocol
 A transaction must get an S lock on a data item before it wishes to READ it
 A transaction must get an X lock on a data item before it wishes to WRITE it
 A TRANSACTION already holding an S lock on a data item can promote
itself to an X lock in order to WRITE it, provided there is no other transaction
also holding an S lock on it already.
 If a lock request is denied, the transaction goes into a WAIT state
 A transaction in a WAIT state resumes operations only when the requested
lock is released by another transaction
 X locks are held until COMMIT/ROLLBACK. S locks are normally the same.
Lock Compatibility Matrix

 A transaction may be granted a lock on an item if


the requested lock is compatible with locks
already held on the item by other transactions
Lock Compatibility Matrix

 Any number of transactions can hold shared locks


on an item
 But if any transaction holds an exclusive on the item
no other transaction may hold any lock on the item.
Lock Compatibility Matrix

 If a lock cannot be granted, the requesting


transaction is made to wait till all incompatible
locks held by other transactions have been
released. The lock is then granted.
Example
 Locking as above is not sufficient to
guarantee serializability if A and B
get updated in-between the read of
A and B, the displayed sum would be
wrong.
 A locking protocol is a set of rules
followed by all transactions while
requesting and releasing locks.
Locking protocols restrict the set of
possible schedules.
Two Phased Locking

Before operating on
any object (a tuple), a
transaction must
acquire a lock on that
object If all transactions obey
Protocol Theorem the “two–phased
locking protocol”, then
all possible interleaved
schedules are
After releasing a lock a serializable
transaction must never
go on to acquire any
more locks.
Lost Update – Example Revisited
Lost Update – Example Revisited

 neither transaction can update balance as both are waiting


for other to release the lock on balance
Relational Integrity – Example Revisited
Relational Integrity – Example Revisited

CORRECT ANSWER
Thank You

You might also like