WEBVTT

1
00:00:07.107 --> 00:00:10.030
In this video, we will discuss Block Storage

2
00:00:10.030 --> 00:00:13.820
and how it compares to File Storage in the Cloud.

3
00:00:13.820 --> 00:00:17.780
Block storage breaks files
into chunks (or blocks) of data

4
00:00:17.780 --> 00:00:21.320
and Stores each block separately under a unique address.

5
00:00:21.325 --> 00:00:27.360
Like direct attached storage and file storage,
block storage also must be attached to a compute

6
00:00:27.360 --> 00:00:30.890
node before it can be utilized for your workloads.

7
00:00:30.890 --> 00:00:36.460
Block storage, like file storage, can be mounted
from remote storage appliances, making it

8
00:00:36.468 --> 00:00:41.880
extremely resilient to failure, and keeping
data far more secure in them, on account of

9
00:00:41.880 --> 00:00:47.310
encryption in transit, and encryption at rest
services, available on these appliances.

10
00:00:47.314 --> 00:00:52.980
Block storage is mounted as a volume to compute
nodes using a dedicated network of fibers,

11
00:00:52.980 --> 00:00:55.580
through which signals move at the speed of
light.

12
00:00:55.588 --> 00:01:00.660
These fiber optic networks are more expensive
to build than the ethernet ones which deliver

13
00:01:00.660 --> 00:01:05.880
File Storage, which is one reason why Block
Storage tends to have a higher price-point.

14
00:01:05.880 --> 00:01:11.400
However, since the traffic is moving faster
and with speed consistency, they are perfect

15
00:01:11.400 --> 00:01:15.000
for workloads that need low-latency storage
to work effectively.

16
00:01:15.000 --> 00:01:20.400
In terms of workloads, it is important to
note that unlike File Storage, which can be

17
00:01:20.400 --> 00:01:26.040
mounted onto 80 compute nodes or more, Block
Storage is normally mounted onto only one

18
00:01:26.040 --> 00:01:27.300
compute node at a time.

19
00:01:27.300 --> 00:01:32.640
Since these disks run at a consistent high
speed, they are perfect for workloads that

20
00:01:32.640 --> 00:01:36.720
need consistently fast storage, such as databases
and mail servers.

21
00:01:36.720 --> 00:01:41.760
Block storage is not suitable for workloads
where there needs to be some level of disk

22
00:01:41.760 --> 00:01:43.620
sharing between compute nodes.

23
00:01:43.620 --> 00:01:49.860
For block storage, as it is for file storage,
you need to take the IOPS capacity of the

24
00:01:49.860 --> 00:01:51.060
storage into account.

25
00:01:51.060 --> 00:01:57.480
Most cloud providers will allow you to specify
IOPS characteristics when you provision storage

26
00:01:57.480 --> 00:02:03.360
and, in some cases, adjust the IOPS of your
storage as you need, so if the requirements

27
00:02:03.360 --> 00:02:08.220
or usage behaviour of an application changes,
you can adjust accordingly.

28
00:02:08.220 --> 00:02:14.220
So, to summarise the commonalities and differences
between these two storage types:

29
00:02:14.220 --> 00:02:19.680
Block and File Storage is taken from appliances
which are maintained by the service provider.

30
00:02:19.680 --> 00:02:25.680
Both are normally highly available and resilient
and will often include data encryption at

31
00:02:25.680 --> 00:02:27.240
rest and in transit.

32
00:02:27.240 --> 00:02:33.180
File storage is attached to a compute node
using an ethernet network, so it is sometimes

33
00:02:33.180 --> 00:02:36.590
called Network attached or NFS Storage.

34
00:02:36.594 --> 00:02:42.920
File storage is very reliable, but the speed
of the connecting network can vary, based on load.

35
00:02:42.925 --> 00:02:49.500
Block storage is attached via a high-speed fiber
network, which is very reliable and consistent.

36
00:02:49.500 --> 00:02:53.520
File storage can be attached to multiple compute
nodes at once.

37
00:02:53.520 --> 00:02:57.180
Block storage can only be attached to one
node at a time.
38
00:02:57.180 --> 00:03:02.880
File storage is a good choice where file shares
are required, where workloads do not require

39
00:03:02.880 --> 00:03:07.560
lightning fast connectivity to storage, or
where cost is a factor.

40
00:03:07.560 --> 00:03:12.680
Block storage is a good choice when supporting
an application that needs consistent fast

41
00:03:12.680 --> 00:03:15.450
access to disk, such as databases.

42
00:03:15.451 --> 00:03:20.280
Remember to consider the IOPS requirements
of the application when provisioning either

43
00:03:20.280 --> 00:03:22.280
file or block storage.

44
00:03:22.285 --> 00:03:24.180
Hey guys, my name is Amy Blea.

45
00:03:24.180 --> 00:03:27.060
I'm on the offering team here with IBM cloud and today

46
00:03:27.060 --> 00:03:31.850
I want to talk to you about some traditional
types of storage: file and block storage.

47
00:03:31.851 --> 00:03:34.380
Kind of give you an overview of what they are

48
00:03:34.380 --> 00:03:38.350
and then give you some tips on how to decide
when to use one over the other.

49
00:03:38.350 --> 00:03:41.020
So, let's get started with block storage.

50
00:03:41.028 --> 00:03:47.040
Block storage is storage where data is
written in raw blocks on the storage and it's
51
00:03:47.040 --> 00:03:52.910
accessed by the servers that you
have through a storage area network.

52
00:03:52.914 --> 00:03:57.240
So, you have all of your servers, they can
either be on the same network with each other or

53
00:03:57.240 --> 00:04:03.390
a different network but they all connect through
the storage, through this storage area network.

54
00:04:03.390 --> 00:04:07.810
Now, some of the advantages of using
block storage is that block storage

55
00:04:07.810 --> 00:04:12.000
provides the lowest possible
latency for your applications,

56
00:04:12.000 --> 00:04:19.770
and you can also use it for those applications
that require high performance or lots of IOPS.

57
00:04:19.770 --> 00:04:24.730
Another thing you'll find with block storage
is that, in general, it is highly redundant.

58
00:04:24.731 --> 00:04:30.840
Most block storage services provide
a capability that's built-in,

59
00:04:30.840 --> 00:04:34.550
where your data is redundant across the volume, so that if,

60
00:04:34.558 --> 00:04:38.280
so there should be
either a volume should go down

61
00:04:38.280 --> 00:04:42.730
or a disk should go away and
you can recover your data

62
00:04:42.730 --> 00:04:47.120
from another place without having any
kind of impact to your application.
63
00:04:47.820 --> 00:04:54.180
Now, file storage is connected to your service
a little differently than block storage is.

64
00:04:54.180 --> 00:05:03.290
Everything is connected on the same network so you
have all of your files or your file share here.

65
00:05:04.620 --> 00:05:09.100
Those are all the files and they
can all be accessed on the same network

66
00:05:09.108 --> 00:05:11.200
by any of the servers on that network.

67
00:05:11.200 --> 00:05:13.620
So, it's network attached storage.

68
00:05:13.622 --> 00:05:19.350
File storage is highly scalable so you can have
multiple file shares on your network

69
00:05:19.350 --> 00:05:22.300
and have all of
your servers attached to that at one time.

70
00:05:22.308 --> 00:05:25.140
It's accessible to multiple runtimes.

71
00:05:25.140 --> 00:05:28.250
So, here in this illustration,
we have a single file share that

72
00:05:28.251 --> 00:05:30.600
has multiple servers accessing it all at once,

73
00:05:30.600 --> 00:05:37.500
and you can also have multiple simultaneous
reads and writes going to your file share all

74
00:05:37.500 --> 00:05:41.870
at one time without having to worry
about your data being overwritten.

75
00:05:42.180 --> 00:05:44.830
So, how do you know when is the
right time to choose

76
00:05:44.830 --> 00:05:48.770
either block storage or file storage
for your application?

77
00:05:48.770 --> 00:05:52.140
Well, first of all, you need to
think about what you're using it for.

78
00:05:52.148 --> 00:05:57.660
So, for example, let's say you have
a VMware configuration where you have

79
00:05:57.660 --> 00:06:03.420
multiple virtual servers with VMware
on them and you need boot volumes.

80
00:06:03.420 --> 00:06:06.750
Well, in that case, you would use block storage.

81
00:06:07.080 --> 00:06:13.380
If you have workloads such as transactional
databases or relational databases that require

82
00:06:13.380 --> 00:06:18.620
very low latency and high performance,
you would choose block storage

83
00:06:19.200 --> 00:06:23.400
In situations where you have a mix of
the structured and unstructured data,

84
00:06:23.405 --> 00:06:28.410
for example, a web hosting server
where you have both, you know,

85
00:06:28.410 --> 00:06:34.110
text files as well as media files,
then you would choose file storage.

86
00:06:34.800 --> 00:06:39.610
And then, if you have a collaborative
space where you need to have multiple

87
00:06:39.611 --> 00:06:44.400
users accessing it all at one time,
working together, doing, you know,

88
00:06:44.400 --> 00:06:49.150
reads and writes all at the same
time, you would choose file storage.

89
00:06:49.860 --> 00:06:53.180
So, like I mentioned in the beginning,

90
00:06:53.180 --> 00:06:56.560
block and file storage are very
traditional types of storage.

91
00:06:56.560 --> 00:07:01.330
They're not as flashy and shiny and
exciting as some of the new

92
00:07:01.330 --> 00:07:04.380
storage services that have entered the market,

93
00:07:04.388 --> 00:07:06.840
but they are still very relevant and useful

94
00:07:06.840 --> 00:07:11.450
for the different types of workloads that you
have, whether it's on premise or in the cloud.

95
00:07:11.450 --> 00:07:16.010
Thanks for checking out this video on the differences
between file and block storage with me today.

96
00:07:16.010 --> 00:07:20.960
If you have any questions or comments or ideas
for other videos, just put something in the

97
00:07:20.960 --> 00:07:25.610
comments below and then if you'd like to see
more videos like this in the future subscribe.

98
00:07:25.610 --> 00:07:29.500
In the next video, we'll start
to look at Object Storage.