0:06	in this presentation we will see about
0:08	the database system architecture we are
0:11	now in part one of the database system
0:14	architecture
0:15	from the last few presentations we
0:17	realized the importance of having dbms
0:19	by organizations because data is crucial
0:23	and the volume of data handled by dbms
0:26	is really fascinating because the data
0:28	that is processed by dbms is not just
0:31	the bit of data or bytes of data they
0:34	range in kilobytes megabytes gigabyte
0:38	terabyte
0:39	petabyte exabyte zettabyte yotabyte and
0:43	whatnot
0:45	in the coming future we may also end up
0:47	in more data volume than what is
0:49	specified on the slide so when a
0:51	database needs to handle this much of
0:54	information obviously the main memory
0:56	cannot handle this though the main
0:58	memory is faster in nature but this main
1:01	memory cannot store the database because
1:03	it's volatile in nature at the same time
1:06	the secondary storage devices like the
1:08	hard disk or the ssds the solid state
1:11	drives will also not be sufficient due
1:13	to the increased volume of data then
1:16	what's the solution for this because
1:18	this contemporary world is handling an
1:20	enormous amount of data so obviously we
1:23	are in the need of storage servers or
1:25	data centers or cloud that can handle
1:28	large volume of data the reason why i am
1:30	projecting this is that i'll give you a
1:32	real time example just see facebook how
1:35	much volume of data it is handling every
1:37	day take another server is a whatsapp
1:39	server which is handling an enormous
1:41	amount of data
1:43	every day every minute every second it
1:45	handles a large volume of data at the
1:48	same time whatever we post it is stored
1:50	in their servers let's say you have
1:52	posted your photo in your facebook
1:54	account and later you have decided that
1:56	you don't want that photo to be posted
1:59	so obviously we will be deleting that
2:01	isn't it so when you delete it's just
2:04	deleted from the view of yours and your
2:06	friends
2:07	is it actually deleted from the facebook
2:09	server
2:10	no it will not so whatever we upload
2:14	everything is stored in the servers and
2:17	think about the complexity of handling
2:19	such high volume of data but still our
2:22
technology and servers are able to
2:24
handle this that's the part of being in
2:26
this contemporary world and coming to
2:28
the dbms architecture this is also
2:31
referred as the database system
2:33
structure
2:34
and what we are going to learn in this
2:36
presentation is the part one of the
2:38
database system architecture at first
2:40
let me show you the architecture diagram
2:43
here is the database system architecture
2:45
if you see the diagram we can see the
2:48
users are at the top level then iv users
2:51
the application programmers the
2:52
sophisticated users and the database
2:55
administrators obviously each users will
2:58
be preferring some set of tools for
3:00
their activities and here is the tool
3:02
level and if you see the architecture
3:04
diagram we have two important components
3:07
the query processor part and the storage
3:09
manager part we will come to this later
3:12
and also if you see the databases are
3:15
actually stored in the disk so in this
3:17
example we are assuming that the
3:19
database is actually stored in the disk
3:21
and that's why this storage is referred
3:23
as the disk storage so in the database
3:26
architecture part 1 and part 2
3:28
presentations we are going to
3:29
exclusively focus on all the individual
3:32
components that constitute this database
3:34
system architecture i mean the
3:36
components in the query processor as
3:38
well as the storage manager in part one
3:41
of database system architecture we are
3:42
going to exclusively focus on the
3:44
storage manager and the disk storage i
3:47
mean we are going to focus on the buffer
3:49
manager the file manager the
3:52
authorization and integrity manager and
3:54
the transaction manager also we are
3:57
going to see about the data indices the
3:59
statistical data and the data dictionary
4:02
so let's see one by one in detail so we
4:05
are clear that we are going to focus on
4:07
the storage manager which is this so we
4:09
are in this part and we are going to
4:11
focus on all these managers before
4:14
seeing the managers one by one let's
4:16
understand why do we need the storage
4:18
manager and what role it plays in the
4:20
database system architecture
4:22
basically this storage manager provides
4:24
the interaction between the low level
4:27
data stored and the application programs
4:29
if you see the architecture diagram
4:32
this storage manager provides the
4:34
interaction between the application
4:36
programs that is from the top level and
4:38
the data stored at the low level so here
4:41
is the actual data that is stored and to
4:44
be precise here is the actual data that
4:46
is stored in the database
4:48
actually this storage manager only
4:50
handles all the queries that are
4:52
submitted to the system what do we mean
4:54
by this if you see the architecture
4:57
diagram the query processor processes
4:59
the queries and this will give the
5:01
request to the next level which is the
5:03
storage manager and that is why the
5:05
third point says that this is also
5:07
having an interaction that is received
5:10
from the query processor to the desk
5:12
storage and also this storage manager is
5:15
having the interaction with the file
5:17
manager you may be asking me a question
5:19
we have already compared file system and
5:21
database system and we understood that
5:24
databases have more advantages over file
5:26
system in effective storing and
5:28
retrieval of data
5:30
even though we are saying that it is a
5:32
database system internally it is made up
5:34
of files only but it's not going to work
5:37
as the regular file system rather it is
5:40
going to work differently why because we
5:43
need to effectively store and retrieve
5:45
the data in database management system
5:47
at the same time the various drawbacks
5:49
that we have seen with the file systems
5:52
needs a solution here though we are
5:54
using internally all files but this is
5:56
not going to work as per the file system
5:58
concept it's going to work as per the
6:01
database system concept and also we know
6:03
that only low level data are stored and
6:06
this low level data is actually the raw
6:08
data that are stored on the disk using
6:10
the file system provided by the
6:12
operating system and obviously whatever
6:15
the interaction that we make see we may
6:17
do some insertion or deletion or
6:19
updation whatever it is coming from the
6:22
front end or from the high level those
6:24
data have to be stored in the database
6:27
to be precise it has to be stored on the
6:29
disk only because this database is on
6:32
the disk only so all the raw data are
6:35
stored on the disk and what format it is
6:37
going to use this is going to use the
6:39
same file system format that is provided
6:42
by the host operating system so we are
6:44
clear that if data are to be stored on
6:46
the disk it uses the file system
6:48
provided by the operating system and
6:50
these raw data are provided from the
6:52
users or from the application layer or
6:55
from the top post layer and what about
6:57
the dml commands what do we mean by dml
7:00
commands these dml statements are the
7:03
statements that deals with the insertion
7:05
of data or deletion of data or updation
7:08
of data or sometimes selection of data
7:11
from the databases we know databases are
7:13
going to store the data and data related
7:16
operations are built by the dml
7:18
statements
7:19
example select insert delete update
7:23
nowadays in the last lecture of this
7:25
chapter we are going to focus on the
7:27
dbms languages at that time you will be
7:29
able to understand what is ddl what is
7:32
dml and other languages also we have a
7:35
separate chapter called sql where we are
7:37
going to exclusively focus on the sql
7:40
the structured query language along with
7:43
ddl dml tcl and dcl operations
7:47
i know things will be unclear but for
7:49
now just understand dml is data related
7:53
so when such commands are received this
7:56
storage manager only translates the
7:58
various dml statements into low-level
8:01
file system commands why because the
8:04
internal organization needs to be
8:06
supported by the operating system and
8:08
this operating system obviously will be
8:10
using some file system and whatever the
8:12
data manipulation language commands that
8:15
we give these things have to be
8:17
translated into the low level file
8:19
system commands in simple terms storage
8:22
manager is responsible for storing
8:24
retrieving and updating the data now we
8:27
will see the various managers that are
8:29
there in the storage manager as per the
8:32
architecture diagram we are sure that
8:34
these are all the four components or
8:36
managers in the storage manager part
8:38
number one the authorization and
8:40
integrity manager number two the
8:42
transaction manager number three the
8:45
file manager and number four the buffer
8:47
manager and also as already stated we
8:50
are going to look at various data
8:52
structures the data files the data
8:54
dictionary and the indices
8:57
we will see one by one now but let's go
8:59
to the architecture diagram at first we
9:02
will focus on the first storage manager
9:04
component the buffer manager and this
9:06
buffer manager is actually responsible
9:08
for fetching the data from the disk
9:11
storage here is the disk storage and
9:13
into the main memory and also this
9:16
buffer manager decides what data to
9:18
cache in the main memory i hope you are
9:21
already aware of cache memory in
9:23
computer systems this memory is mainly
9:26
used for faster access
9:28
so buffer manager takes cares of this at
9:31
the same time this buffer manager is
9:33
very crucial in terms of another
9:35
important operation that is handling the
9:37
size of the data we know the databases
9:40
will be larger than our actual main
9:42
memory so this buffer manager will also
9:44
take care of this memory management and
9:47
coming to the second one the file
9:49
manager we know already this is mainly
9:52
for allocating the space on the disk
9:54
storage we know actually the databases
9:56
are already stored on the disk and we
9:58
may be inserting new data or deleting
10:00
the existing data or updating the
10:02
existing data this file manager is
10:05
responsible for allocating the space on
10:07
the disk storage and also the data
10:10
structures we have a lot of data
10:12
structures followed in the disk storage
10:14
and this file manager is used to
10:16
represent the information stored on the
10:18
disk
10:19
and coming to the next one the
10:20
authorization and integrity manager this
10:23
component or this manager is used to
10:25
test for the satisfaction of all the
10:28
integrity constraints that we enforce on
10:30
the database say for example balance
10:32
should not be less than zero in such
10:34
case if any operation that is carried
10:36
out on the database where the balance is
10:39
going to be zero or lesser than zero
10:41
that operation needs to be aborted that
10:43
should not be permitted because we have
10:45
given an integrity constraint that
10:46
balance should not be less than zero so
10:49
all these integrity constraints are
10:51
handled by this authorization and
10:53
integrity manager at the same time
10:55
authorization of data i mean who can
10:58
access what that is also actually
11:00
handled by this authorization and
11:02
integrity manager and coming to the last
11:04
manager the transaction manager and this
11:07
is the very important part because this
11:10
ensures that the database remains
11:12
consistent despite system failures or
11:15
any hardware or software failures at the
11:17
same time this transaction manager
11:20
ensures that the concurrent operations
11:22
carried out on this database are
11:25
actually not conflicting with each other
11:27
i mean concurrent operations are
11:29
permitted multiple users can access the
11:32
shared data item at the same time
11:34
without conflicting each other so this
11:37
is ensured by the transaction manager
11:40
anyway in the next lecture which is the
11:42
part 2 of the database system
11:44
architecture i will elaborate more on
11:46
this transaction manager we are done
11:48
with the storage managers what are all
11:50
the components that we have seen
11:52
we have seen about the authorization and
11:54
integrity manager the transaction
11:56
manager file manager and buffer manager
11:59
before we sign out let's see the various
12:01
data structures used what are all the
12:03
data structures we are going to focus on
12:05
the data files actually this stores the
12:08
database itself here is the data files
12:11
so here is the actual database that is
12:13
residing
12:14
and coming to the next one which is the
12:16
data dictionary we have already seen
12:19
about metadata that is data about other
12:22
data that is stored so these metadata
12:24
information are stored by the data
12:26
dictionary and coming to the third data
12:29
structure which is indices we know
12:31
obviously data are stored in the
12:32
databases and we need to ensure that the
12:35
data are accessed fast
12:37
storage and retrieval are efficient and
12:39
effective and we are also sure that
12:41
databases are handling large volume of
12:44
data in such cases referring to the
12:47
exact data item that we want say our
12:49
data may be in the first location or it
12:51
may be in the last location but it
12:53
should not be the case that databases
12:55
takes a lot of time to process our
12:57
request it should be as fast as it can
13:00
so in such case indices helps us to
13:03
retrieve the data faster anyway in the
13:06
chapter indexing and hashing i will be
13:08
talking about indices elaborately for
13:10
now just have this this is used to
13:12
provide fast access to our data item and
13:15
if you want an analogy this is like
13:18	index pages in our textbook why do we
13:20	need index pages in our textbook suppose
13:23	we want to refer something we need not
13:25	go sequentially to refer that if you go
13:27	to the index file this index file will
13:29	take us to the exact location what we
13:31	are looking for and the similar concept
13:33	is applied on the databases for faster
13:36	access so what we have seen now we have
13:38	seen the various data structures like
13:40	the data files the data dictionary and
13:43	the indices one thing we have left
13:45	what's that the statistical data
13:48	so this part is mainly storing the
13:50	statistics about the data so this will
13:53	also help us to take some decisions for
13:55	managing the databases
13:57	so what we have learned today in part
14:00	one of database system architecture we
14:02	focused on the storage manager and the
14:04	disk storage and already we have seen
14:06	about the various types of users in one
14:09	of our previous presentations so in the
14:11	next presentation we will focus on the
14:13	remaining part which is the query
14:15	processor
14:17	so i hope the session is informative and
14:19	thank you for watching
14:21	[Music]
14:22	[Applause]
14:24	[Music]
14:32	you