This presentation will give you a quick introduction to Docker & how containers are built, run & shared.
Transcript:
00:02
welcome to this Alan talks tech
00:04
presentation on containers and docker if
00:07
you would like more information on
00:08
spirent and how to test please visit our
00:11
website at wwc are calm if you would
00:15
like to view other alan talk tech
00:17
presentations simply search on youtube
00:20
for alan talks tech if you would like to
00:23
get a copy of this presentation along
00:24
with many others please visit my wiki at
00:27
Alan test wiki pbworks com now let's
00:36
take a look at what many believe is
00:39
going to be the next step in network
00:41
functions virtualization or NFV of
00:49
course we are all familiar with shipping
00:52
containers these are being used for many
00:54
years as a very efficient and standard
00:56
means of shipping our goods all over the
00:59
planet a few years ago when i was giving
01:06
presentations on how data centers were
01:08
being put together i described one novel
01:10
concept we're shipping containers were
01:13
pre-configured with all the hardware
01:15
required to build a data center in fact
01:18
each container was a mini data center in
01:21
its own right all that was required was
01:24
for the container to be hooked up to
01:26
power cooling and communications and the
01:29
system was ready to go multiple
01:32
containers could simply be racked and
01:34
stacked to create very large data
01:36
centers such as this microsoft data
01:39
center in North Lake Illinois today we
01:47
are using containers again to help build
01:49
the data center but these are software
01:52
containers not physical shipping
01:54
containers each container is very much
01:56
like a virtual machine containing the
01:59
applications OS requirements and all of
02:02
the operating dependencies
02:08
so let's take a look at how virtual
02:11
machines and containers stack up
02:13
alongside each other VMS require a
02:17
physical server or metal to run on they
02:20
also require a hypervisor the VMS run on
02:24
top the hypervisor each VM will then
02:26
need an operating system to support the
02:29
applications on the other hand
02:33
containers only need the physical server
02:35
or metal and one instance of an
02:38
operating system we do not need to load
02:41
an operating system for each container
02:43
as we would for a virtual machine the
02:46
container does not need a vm as each
02:48
container is sharing the common OS
02:51
kernel overall the containers are much
02:54
more lightweight as only one operating
02:57
system typically a very stripped down
02:59
linux operating system utilizing a
03:03
system such as core OS in addition as no
03:08
vm is required far fewer system
03:10
resources are needed and the container
03:12
along with its applications will deploy
03:15
much more rapidly the OS is already
03:19
running as a rule of thumb many people
03:21
believe that up to six containers or six
03:24
times as many containers may be run on
03:27
any given platform or metal compared to
03:30
virtual machines
03:35
dr. is built on top of linux containers
03:39
or lxc and adds image management and
03:42
deployment assistance for virtualizing
03:44
applications docker provides for
03:47
automation and rapid revision of lenox
03:50
container groups without requiring a
03:52
virtual machine docker provides an API
03:55
and applications programmable interface
03:58
that extends the functionality of alexey
04:01
or linux containers for building
04:03
platform as a service offerings darker
04:07
today is one of the most popular methods
04:09
for building running managing and
04:12
distributing containers there are other
04:15
packages available for helping you build
04:17
a container but as I said by far docker
04:20
is way ahead of most of the other
04:22
competing systems
04:28
let's take a look at how you can build
04:31
and share a container firstly docker is
04:34
a client-server application building a
04:38
container to run an application is a
04:40
little like making a cake first we need
04:43
to start with a recipe this is called
04:46
the dockerfile the dockerfile is a
04:48
default name given to the file that
04:51
docker will use to build the image the
04:54
image in turn will then provide
04:56
everything that is required to run the
04:58
container the image may be large but the
05:01
actual running container holding
05:03
application and its dependencies may be
05:05
very light only the required components
05:09
all dependencies from the image will be
05:11
used to run the container remember the
05:14
basic OS is already running and will be
05:17
used by the container this is one of the
05:20
key reasons that containers are so fast
05:22
to spin up containers do not need to
05:24
start relatively complex of em and then
05:27
wait for the OS to boot up so when
05:31
building a docker file for example to
05:34
run an Apache server you would add the
05:37
Apache code or tell the file where to
05:40
get the code from and some information
05:43
on where to get a bunt to from add your
05:46
HTML files or point to where the HTML
05:49
files are then using your docker client
05:52
you can tell the docker host to build
05:55
the image you can then run the image as
05:58
a container if you're happy with the
06:00
image you can then push the image up
06:03
into the docker registry as I said
06:06
docker contains two components the
06:09
daemon runs on a host machine and does
06:11
the heavy lifting of building running
06:13
and distributing docker containers the
06:16
client is a docker binary that accepts
06:19
commands from the user and communicates
06:21
back and forth with the daemon for many
06:25
of us when we're building the container
06:27
will often be using the same PC so the
06:30
client and the host will be all residing
06:32
on the same machine but it's also
06:34
possible and it gives us a lot of
06:36
flexibility that the daemon can run up
06:38
in
06:38
cloud so from my pc i can run the client
06:41
but i could build the container build
06:43
the images up in the cloud and of course
06:45
i can store those images which are used
06:49
to build the containers up in the docker
06:51
registry which is effectively up in the
06:53
cloud using the docker registry to build
06:59
your system once you've built all of
07:01
your containers with all of the required
07:03
applications each container can be a
07:07
separate application all of the
07:09
containers or applications can be stored
07:11
in the registry and pull down and
07:13
deployed on our server as required of
07:16
course others may also share these
07:19
containers with their pre-configured
07:21
applications and dependencies making
07:24
containers very portable across two
07:26
different operating environments the
07:29
registry can either be a public or
07:31
private registry either on the docker
07:33
cloud or on a private corporate server I
07:41
hope you found this presentation on
07:44
containers and docker useful again if
07:47
you'd like to get additional information
07:48
on spirent and what our testing
07:50
capabilities are please visit our
07:52
website at wwc are calm you can also of
07:57
course look at other YouTube videos by
07:59
searching on Allen talks tech and if
08:02
you'd like a copy of this presentation
08:03
along with many other more detailed
08:06
presentations please visit my wiki at
08:08
Allen test wiki pbworks com once again
08:14
thank you for viewing