1
00:00:06,930 --> 00:00:09,373
- Now, let's talk about Amazon Aurora.

2
00:00:10,920 --> 00:00:14,810
With Amazon Aurora, we
gain access to a database

3
00:00:14,810 --> 00:00:17,070
that is compatible with,
that is essentially

4
00:00:17,070 --> 00:00:22,070
a drop-in replacement for
MySQL 5.6 or Postgres 9.6.

5
00:00:24,300 --> 00:00:29,050
And so, for most used cases,
not all but most used cases,

6
00:00:29,050 --> 00:00:32,960
you can see companies have
seen three to five times

7
00:00:32,960 --> 00:00:36,610
performance increase with
no changes to the schema

8
00:00:36,610 --> 00:00:38,270
and no changes to code.

9
00:00:38,270 --> 00:00:41,140
Now, one of the great things about Aurora

10
00:00:41,140 --> 00:00:43,860
besides being a very
performant replacement

11
00:00:43,860 --> 00:00:48,860
for MySQL or Postgres is
that we get a storage plane

12
00:00:49,920 --> 00:00:53,310
up to 64 terabytes that is auto-scaled

13
00:00:53,310 --> 00:00:56,330
and inherently replicated, right?

14
00:00:56,330 --> 00:00:58,010
Now, if you were to compare this.

15
00:00:58,010 --> 00:01:03,010
If two RDS, let's say
running Postgres on RDS.

16
00:01:03,080 --> 00:01:08,080
With RDS, if you wanted a
larger storage platform,

17
00:01:08,550 --> 00:01:12,450
you would have to create an
entirely new database instance

18
00:01:12,450 --> 00:01:16,020
to gain access to a
larger underlying storage.

19
00:01:16,020 --> 00:01:19,310
But with Amazon Aurora, the storage

20
00:01:19,310 --> 00:01:24,150
just grows automatically and
it grows in certain increments.

21
00:01:24,150 --> 00:01:27,260
As your dataset grows, the
underlying storage volume

22
00:01:27,260 --> 00:01:30,160
will automatically grow to
accommodate that dataset.

23
00:01:30,160 --> 00:01:31,290
There's nothing for you to do.

24
00:01:31,290 --> 00:01:34,860
You don't have to pre-provision
a certain amount of storage.

25
00:01:34,860 --> 00:01:38,810
It just automatically
grows up to 64 terabytes.

26
00:01:38,810 --> 00:01:42,550
And there's nothing we
need to do with Aurora

27
00:01:42,550 --> 00:01:45,720
in order to ensure that
we have multiple copies

28
00:01:45,720 --> 00:01:46,553
that are live.

29
00:01:46,553 --> 00:01:49,870
It automatically maintains numerous copies

30
00:01:49,870 --> 00:01:53,260
across numerous availability zones, right?

31
00:01:53,260 --> 00:01:56,990
So a lot of that high
availability fault tolerance

32
00:01:56,990 --> 00:02:01,990
and data durability, it's already
built-in to Amazon Aurora.

33
00:02:02,400 --> 00:02:05,930
We also get up to 15 read replicas,

34
00:02:05,930 --> 00:02:08,300
very easy read replica creation.

35
00:02:08,300 --> 00:02:09,920
And then we also gain access

36
00:02:09,920 --> 00:02:12,430
to what we call Aurora Multi-Master

37
00:02:12,430 --> 00:02:16,290
where we can have multiple
masters to which we can write to.

38
00:02:16,290 --> 00:02:18,590
Not only can we have
those 15 read replicas

39
00:02:18,590 --> 00:02:21,800
but we can also write to multiple masters

40
00:02:21,800 --> 00:02:26,020
and those are also the
replication between those two

41
00:02:27,070 --> 00:02:30,590
is managed by Amazon,
by the Aurora service.

42
00:02:30,590 --> 00:02:34,820
And lastly, we also
have Aurora Serverless.

43
00:02:34,820 --> 00:02:38,960
Now, with normal Aurora, even
though Amazon is managing

44
00:02:38,960 --> 00:02:41,490
all of the underlying infrastructure.

45
00:02:41,490 --> 00:02:46,080
The storage plane, the compute
service and everything else.

46
00:02:46,080 --> 00:02:48,240
There's still something for us to manage,

47
00:02:48,240 --> 00:02:51,690
there's still an instance of
Aurora that we need to create

48
00:02:51,690 --> 00:02:53,960
and modify options on.

49
00:02:53,960 --> 00:02:56,763
But with Aurora Serverless, there isn't.

50
00:02:56,763 --> 00:03:00,430
There's nothing there
until we actually need it.

51
00:03:00,430 --> 00:03:02,930
So Aurora Serverless
is more like on demand

52
00:03:02,930 --> 00:03:07,230
when our application needs
that database engine,

53
00:03:07,230 --> 00:03:09,830
it boots up, it makes itself available,

54
00:03:09,830 --> 00:03:11,690
we do our writes, we do our reads

55
00:03:11,690 --> 00:03:15,080
and then we don't need
it anymore, it goes away.

56
00:03:15,080 --> 00:03:17,500
So we have a lot of really and of course,

57
00:03:17,500 --> 00:03:20,957
we'll take a look a closer
look at Aurora Serverless

58
00:03:22,890 --> 00:03:24,730
in an upcoming section.

59
00:03:24,730 --> 00:03:27,080
But of course, remember the key point here

60
00:03:27,080 --> 00:03:30,690
is that Amazon Aurora
can be a very performant

61
00:03:30,690 --> 00:03:34,253
drop-in replacement for MySQL or Postgres.