Would you like to inspect the original subtitles? These are the user uploaded subtitles that are being translated:
1
00:00:01,390 --> 00:00:03,220
In this lesson,
we want to examine
2
00:00:03,220 --> 00:00:07,240
the concept of Real
Application Clusters, or RAC,
3
00:00:07,240 --> 00:00:08,770
in an Oracle database.
4
00:00:08,770 --> 00:00:11,800
RAC is a hot topic today
in the Oracle World
5
00:00:11,800 --> 00:00:15,130
because RAC is Oracle's
clustering solution.
6
00:00:15,130 --> 00:00:17,470
And the concept of
clustering involves
7
00:00:17,470 --> 00:00:20,470
taking multiple
computers and having them
8
00:00:20,470 --> 00:00:23,410
act in concert together as one.
9
00:00:23,410 --> 00:00:27,580
RAC is a pay-extra option
for the Oracle database.
10
00:00:27,580 --> 00:00:31,090
So you can have an Oracle
database Enterprise license
11
00:00:31,090 --> 00:00:34,480
and yet not be licensed to
run RAC, or Real Application
12
00:00:34,480 --> 00:00:35,350
Clusters.
13
00:00:35,350 --> 00:00:37,540
It is a pay-extra option.
14
00:00:37,540 --> 00:00:39,260
It is rather expensive.
15
00:00:39,260 --> 00:00:42,040
But that is because
it affords a company
16
00:00:42,040 --> 00:00:45,220
a lot of positive
possibilities when
17
00:00:45,220 --> 00:00:47,110
it comes to their database.
18
00:00:47,110 --> 00:00:50,230
First of all, it's a high
availability solution.
19
00:00:50,230 --> 00:00:53,800
Because in RAC the
software allows machines
20
00:00:53,800 --> 00:00:59,500
to work together, if any one
machine is taken down or fails
21
00:00:59,500 --> 00:01:02,800
or has to even be patched,
the other machines
22
00:01:02,800 --> 00:01:04,090
continue the work.
23
00:01:04,090 --> 00:01:08,120
So it's high availability
for a company's applications.
24
00:01:08,120 --> 00:01:11,510
The next is that it's
a scalability solution.
25
00:01:11,510 --> 00:01:13,900
So if a company
wants to grow, they
26
00:01:13,900 --> 00:01:16,600
have the option in Real
Application Clusters
27
00:01:16,600 --> 00:01:19,960
of simply adding another
machine to the cluster.
28
00:01:19,960 --> 00:01:25,120
So let's use an example for
why a company would want to use
29
00:01:25,120 --> 00:01:26,740
Real Application Clusters.
30
00:01:26,740 --> 00:01:29,290
Let's say that we
have a small company.
31
00:01:29,290 --> 00:01:35,140
We've started a company, and it
does online ordering for books.
32
00:01:35,140 --> 00:01:37,510
Initially, our customer
base, let's say,
33
00:01:37,510 --> 00:01:39,670
is about 100 customers.
34
00:01:39,670 --> 00:01:43,150
And we have an employee
base of about five people.
35
00:01:43,150 --> 00:01:45,220
In that situation,
we would probably
36
00:01:45,220 --> 00:01:48,050
need a fairly small
database server.
37
00:01:48,050 --> 00:01:51,400
We need it to store things
like basically our ordering,
38
00:01:51,400 --> 00:01:54,430
the different orders that
are coming into our company,
39
00:01:54,430 --> 00:01:57,280
and also maybe
employee information.
40
00:01:57,280 --> 00:02:00,670
So we settle on a
small to medium sized
41
00:02:00,670 --> 00:02:04,810
commercial hardware server,
put the Oracle database on it,
42
00:02:04,810 --> 00:02:06,940
and the company begins to grow.
43
00:02:06,940 --> 00:02:11,020
And eventually, the customer
base grows to about 10,000.
44
00:02:11,020 --> 00:02:14,410
And our small computer that we
had when our company started
45
00:02:14,410 --> 00:02:16,150
is no longer viable.
46
00:02:16,150 --> 00:02:18,070
So our only option--
47
00:02:18,070 --> 00:02:21,420
without RAC-- is to get
rid of the old server,
48
00:02:21,420 --> 00:02:27,460
buy a new server, maybe a 4U
higher end commercial server
49
00:02:27,460 --> 00:02:29,320
and run our Oracle
database on that.
50
00:02:29,320 --> 00:02:31,050
Well, then, our company
grows even more.
51
00:02:31,050 --> 00:02:33,650
And let's say our customer
base is about 250,000,
52
00:02:33,650 --> 00:02:35,020
a quarter million.
53
00:02:35,020 --> 00:02:39,310
Now our second server that
we had is inadequate as well.
54
00:02:39,310 --> 00:02:41,890
And so now, we have to go
out and buy another server.
55
00:02:41,890 --> 00:02:44,620
So without the
option of RAC, we can
56
00:02:44,620 --> 00:02:47,080
spend quite a bit of
money trying to keep pace
57
00:02:47,080 --> 00:02:48,640
with the growth of our company.
58
00:02:48,640 --> 00:02:52,330
But with RAC, we can simply
have a scalability solution
59
00:02:52,330 --> 00:02:55,090
that allows us to scale
our database system
60
00:02:55,090 --> 00:02:59,650
upwards by simply adding
machines to the cluster.
61
00:02:59,650 --> 00:03:03,700
So to differentiate RAC from
the more common single instance
62
00:03:03,700 --> 00:03:05,500
architecture, let's
look a little bit
63
00:03:05,500 --> 00:03:07,360
about how that works.
64
00:03:07,360 --> 00:03:09,850
So in a single
instance architecture,
65
00:03:09,850 --> 00:03:13,340
the database operates
entirely on a single machine.
66
00:03:13,340 --> 00:03:14,920
So this is a machine.
67
00:03:14,920 --> 00:03:17,110
And it's composed
of the instance,
68
00:03:17,110 --> 00:03:19,000
which is the background
processes and memory
69
00:03:19,000 --> 00:03:21,130
caches, and the disk.
70
00:03:21,130 --> 00:03:23,370
So the processes
run on this server.
71
00:03:23,370 --> 00:03:26,570
The caches are stored in
RAM that's on this server.
72
00:03:26,570 --> 00:03:28,930
And the disk is on
this server as well.
73
00:03:28,930 --> 00:03:32,650
So this is a single instance
architecture-- very common.
74
00:03:32,650 --> 00:03:35,050
Works fine for
many applications.
75
00:03:35,050 --> 00:03:38,650
But it doesn't operate
as a scalability solution
76
00:03:38,650 --> 00:03:40,300
or a high availability solution.
77
00:03:40,300 --> 00:03:44,110
If this machine's power system,
for instance, were to go out,
78
00:03:44,110 --> 00:03:46,450
your data is inaccessible--
79
00:03:46,450 --> 00:03:48,820
processes and caches
being the instance,
80
00:03:48,820 --> 00:03:51,530
and the disk being the database.
81
00:03:51,530 --> 00:03:55,530
But RAC is a multi-instance
architecture.
82
00:03:55,530 --> 00:03:59,570
So in RAC, we may have
three physical machines.
83
00:03:59,570 --> 00:04:02,030
They are three different
physical machines.
84
00:04:02,030 --> 00:04:05,180
They each have their own
instance, their processes
85
00:04:05,180 --> 00:04:06,560
and caches.
86
00:04:06,560 --> 00:04:10,760
But the disk, or the database,
portion of the architecture
87
00:04:10,760 --> 00:04:11,810
is separate.
88
00:04:11,810 --> 00:04:15,350
So we might actually have
a separate disk unit.
89
00:04:15,350 --> 00:04:17,270
And the most important
thing about the disk
90
00:04:17,270 --> 00:04:21,350
is that it's shared
between all three machines.
91
00:04:21,350 --> 00:04:24,920
So the disk has to be
available to every one
92
00:04:24,920 --> 00:04:26,720
of these separate machines.
93
00:04:26,720 --> 00:04:29,840
And shared disks can be done
a number of different ways,
94
00:04:29,840 --> 00:04:32,600
like storage area
networks or network
95
00:04:32,600 --> 00:04:34,370
attached storage, a NAS.
96
00:04:34,370 --> 00:04:38,090
And it also has to have a
network-aware file system,
97
00:04:38,090 --> 00:04:40,310
or a global file system--
98
00:04:40,310 --> 00:04:44,510
something like NFS, the
Network File System, or ASM
99
00:04:44,510 --> 00:04:47,510
in Oracle, Automatic
Storage Management.
100
00:04:47,510 --> 00:04:51,780
So there is a separation between
the instance and the database.
101
00:04:51,780 --> 00:04:56,400
So each physical machine will
be connected to that storage.
102
00:04:56,400 --> 00:04:58,500
But that alone isn't enough.
103
00:04:58,500 --> 00:05:02,370
Because in RAC, one of the
advances that came out in 9i
104
00:05:02,370 --> 00:05:06,810
was the ability to be able to
let each one of these servers
105
00:05:06,810 --> 00:05:08,430
talk to another.
106
00:05:08,430 --> 00:05:10,850
The instances are shared.
107
00:05:10,850 --> 00:05:12,390
They're separate,
but they're also
108
00:05:12,390 --> 00:05:15,990
shareable through a high-speed
network interconnect,
109
00:05:15,990 --> 00:05:18,480
so something like
a high-speed switch
110
00:05:18,480 --> 00:05:21,270
that will connect all three
of those physical machines
111
00:05:21,270 --> 00:05:21,990
together.
112
00:05:21,990 --> 00:05:24,930
Thus, if any one of
the machines is lost,
113
00:05:24,930 --> 00:05:27,270
processing continues
on the other two.
114
00:05:27,270 --> 00:05:30,490
If two machines are lost,
processing continues.
115
00:05:30,490 --> 00:05:33,570
So this is important
for high availability.
116
00:05:33,570 --> 00:05:36,510
But it's also important with
something like patching.
117
00:05:36,510 --> 00:05:39,780
So if you need to do an
operating system patch on one
118
00:05:39,780 --> 00:05:42,990
of the physical machines, it
needs to be generally taken
119
00:05:42,990 --> 00:05:44,430
down or rebooted.
120
00:05:44,430 --> 00:05:47,790
With RAC, we can do
that and still continue
121
00:05:47,790 --> 00:05:49,770
database processing.
122
00:05:49,770 --> 00:05:53,490
So each one of these machines
has a separate instance that
123
00:05:53,490 --> 00:05:56,070
connects to shared storage.
9900
Can't find what you're looking for?
Get subtitles in any language from opensubtitles.com, and translate them here.