What is concurrency control in database transactions? Define concurrency control, Why do we need concurrency control schemes? concurrent transactions
Concurrency Control in DBMS
Definition:
The process of executing many
different simultaneous (concurrent execution) transactions by complying with ACID
properties in a multi-user database system is called as concurrency control.
“Why concurrent execution is needed?”
is discussed in this post. Please refer.
Discussion:
Executing transactions in serial
order, ie., one after the other maintains the integrity of data. All
transactions that are executed serially will show a safe, consistent, durable
database. But the major disadvantage is the time. Executing individual transactions
in serial order consumes lot of time and reduces the transaction throughput. Hence,
we need to execute multiple transactions simultaneously.
Assume that you are executing two
transactions T1 and T2 simultaneously. Also assume that T1 transfers money from
account A to account B and T2 withdraws money from account C. If we execute T1
and T2 simultaneously, the execution will not harm the consistency of the database
because both transactions are working on different data items.
Let us suppose T1 transfers money
from account A to account B and T2 withdraws money from account B. What would
happen if we execute T1 and T2 simultaneously? As both transactions are working
(writing) on the same data items (in this case, both T1 and T2 changes the
value of data item B), whichever writes later will be the final value of B
which is not correct.
From the above discussion we would
understand that, the simultaneous transactions have to be handled with extra
care to ensure the integrity of the data and the database if they read or write
same data items. The process that is
used to ensure this type of safe execution of transaction is called concurrency
control.
********************
Go to Transaction Management in DBMS home page