PARALLEL

DATABASES

By: Sgt Salum Seif Said, Email: seif@skt.ac.tz

What is Parallel Database?

• A parallel database system seeks to improve

performance through parallelization of various
operations like loading data, building index and
evaluating queries.
• Parallel databases improve processing and
input/output speeds by using multiple CPUs and
disks in parallel.
• The operations are performed simultaneously, as
opposed to serial processing.
• Centralized and client–server database systems are
not powerful enough to handle such applications.
Goals of Parallel Database

• Improve Performance
• Improve availability of data:
• Improve reliability:
• Provide distributed access of data
Parameters for Parallel
Databases
• Response time:
It is the time taken to complete a single
task for given time.
• Speed up in Parallel database:
Speed up is the process of increasing
degree of (resources) parallelism to
complete a running task in less time.
The time required for running task is
inversely proportional to number of
resources.
Parameters for Parallel
Databases-Speed Up
Speedup = Time1 / Timem
Time1 is the time it takes to
execute a task using only one
processor, while

Timem is the time it takes to

execute that same task using m
processors.
Parameters for Parallel
Databases - Speedup
Parameters for Parallel
Databases - Speedup
• Speedup can be Linear or sub-linear.
Linear Speedup:
• Speedup is linear if the speedup is N.
• That is, the small system elapsed time is
N times larger than the large system
elapsed time (N is number of resources
say CPU)
Parameters for Parallel
Databases - Speedup

• Linear speed-
up is N (Number of
resources)
• Speed-up is
sub-linear if speed-
up is less than N.
Parameters for Parallel
Databases - Speedup
• If the Speedup is N. i.e. Linear, then it means
the expected performance is achieved.
• If the Speedup is not equal to N, then
following two cases possible:
• Case 1: If Speedup > N, then it means the
system performs more than it designed for. The
Speedup value in this case would be less than 1.
• Case 2: If Speedup < N, then it is Sub-linear. In
this case, the denominator (large system elapsed
time) is more than the single machine’s elapsed
time.
Parameters for Parallel
Databases – Scale up
• Scale-up is the ability to keep performance
constant, when number of process and
resources increases proportionally.
Parameters for Parallel
Databases – Scale up
• For example

• A 4-processor system can provide the

same response time with a workload of 400
transactions per minute as the response
time of a single-processor system that
supports a workload of 100 transactions
per minute.
Parameters for Parallel
Databases – Scale up
• Scaleup is calculated using the following
formula:
Scaleup = Volumm / Volum1
• Volumm = is the transaction volume
executed in a given amount of time using
m processors.
• Volum1 = is the transaction volume
executed in the same time using one
processor.
Parameters for Parallel
Databases – Scale up
• For our previous example:
• Scaleup = 400 / 100
• Scaleup = 4
• This scaleup of 4 is achieved with 4
processors. This is an example of ideal
(linear) scaleup.
Parallel Database
Architecture

• Shared memory system

• Shared Disk System
• Shared nothing disk system
Parallel Database Architecture
- Shared memory system

• Shared memory system uses

multiple processors which is attached
to a global shared memory via
intercommunication channel or
communication bus.
Parallel Database Architecture
- Shared memory system
Parallel Database Architecture
- Shared memory system
Advantages of shared memory system
 Data is easily accessible to any processor.
 One processor can send message to other efficiently.
Disadvantages of Shared memory system
 Waiting time of processors is increased due to more
number of processors.
 Bandwidth problem. (Bandwidth describes the
maximum data transfer rate of a network or Internet
connection.)
Parallel Database Architecture
- Shared Disk system
• Shared disk system uses multiple processors which
are accessible to multiple disks via
intercommunication channel and every processor
has local memory.
• Each processor has Its own memory so the data
sharing is efficient. The system built around this
system are called as clusters.
Parallel Database Architecture
- Shared Disk system
Parallel Database Architecture
- Shared Disk system
Advantages of Shared Disk System
• Fault tolerance is achieved using shared disk system.
• Fault tolerance: If a processor or its memory fails, the
other processor can complete the task. This is called
as fault tolerance.
Disadvantage of Shared Disk System
• Shared disk system has limited scalability as large
amount of data travels through the interconnection
channel.
• If more processors are added the existing processors
are slowed down.
Parallel Database Architecture
- Shared nothing disk system
• Each processor in the shared nothing system
has its own local memory and local disk.

• Processors can communicate with each other

through intercommunication channel.

• Any processor can act as a server to serve the

data which is stored on local disk.
Parallel Database Architecture
- Shared nothing disk system
Parallel Database Architecture
- Shared nothing disk system
Advantages of Shared nothing disk system
• Number of processors and disk can be connected as
per the requirement in share nothing disk system.
• Shared nothing disk system can support for many
processor, which makes the system more scalable.
Disadvantages of Shared nothing disk system
• Data partitioning is required in shared nothing disk
system.
• Cost of communication for accessing local disk is
much higher.
Parallel Query Processing

• Different queries or transaction can be

done in parallel with each other.
Query Parallelism:
• Executing database query/queries in
parallel.
• The concept of parallelism can be
exploited in executing multiple
database queries in parallel.
Parallel Query Processing
Techniques of query Evaluation

• There are two techniques used in query

evaluation are as follows: -
1. Inter Query Parallelism
2. Intra Query Parallelism