Identify Data Center Components

0:00Welcome to Identify Data Center Components.

0:03You know, when I was very early on in my career,

0:05I definitely thought the Data Center was really

0:07all about the network, right?

0:09Like, we've got a primary data center.

0:10That's the network core.

0:11We have a backup data center in case the first data center goes

0:14down, and then we reroute all of our network traffic

0:17through there.

0:18But again, that was just me as a junior network

0:22engineer where my entire focus was on the networking.

0:26Well, the reality is that there's a whole lot more

0:28going on inside the data center than just the networking.

0:31And even though we do need to understand the networking,

0:34we're going to take a look at all the other components that

0:35also exist inside a data center, because really there

0:38are two main things that businesses care about

0:40inside these data centers--

0:41[INAUDIBLE] the whole reason we have them, right?

0:43Number 1, the applications-- whatever

0:46applications we use to conduct our business,

0:48we want those to stay online.

0:50And number 2 would be the data.

0:51We store a lot of data, whether it's customer information

0:54or well-kept secrets.

0:56You got company IP, Intellectual Property.

0:59We store a lot of data inside our data centers as well.

1:02How we get there from a business perspective--

1:04they don't care about that.

1:05They leave that to us.

1:07But hey, you know what?

1:08If we're going to help manage and maintain

1:10and troubleshoot a data center, then we

1:11need to know all of the components.

1:13So stick with me throughout the course of this entire skill,

1:16because we're going to have a lot of fun

1:18going through all of the different components of a data

1:21center.

1:21I hope this has been informative for you,

1:23and I'd like to thank you for viewing.

0:00It's hard to connect to anything without a network.

0:02[MUSIC PLAYING]

0:05By the end of our time here together,

0:07we will be able to identify the primary components

0:09inside a data center network.

0:11And as I said in the intro, the primary purpose

0:13of a data center is not to just simply have a network.

0:16It's to connect things.

0:17Of course, we still have to have that network in order

0:19to connect those things.

0:21So let's take a look at what a data center network looks

0:24like from the ground-up.

0:25Let's start where we connect the servers.

0:27So let's say we have a few servers here,

0:29and we'll draw them out like this.

0:30And these servers need to connect into something.

0:32What are we connecting these into?

0:34Well, we're going to connect them into a switch,

0:37and this switch is typically called an access switch.

0:40And we call them this, because access switches

0:41provide access to whatever we're connecting them to.

0:45These are typically installed into the top of the racks.

0:47We call these a Top of Rack switch, or ToR,

0:50because they go into the top of the same rack

0:53that these servers right here are installed into.

0:55That makes these cables right here a little bit shorter

0:58and more manageable for us.

1:00However, we could also go with an End of Row or Middle

1:02of Row methodology.

1:03And what this means is if we have a row of racks--

1:06let's say we just have, inside of a data center, about four

1:09or five racks in a row, we might install the access switch here

1:13at the end of the row, or here in the middle of the row.

1:15The middle of the row probably makes a little bit more sense,

1:18because what that means is all of the servers in all

1:20these other racks have to be pulled

1:22all the way across to that Middle of Row switch.

1:25Alternatively, if it's at the end of the row,

1:27then the servers all the way at the other end

1:29have to be pulled pretty far to get to that End of Row switch.

1:32We would typically have a lot of access switches in the data

1:35center, so let's try another one here.

1:37And this one can connect to other servers--

1:39probably in the same rack if it's a Top of Rack,

1:42otherwise End of Row or Middle of Row options work.

1:45And these switches then connect to other switches.

1:48Now, depending on the size of the environment,

1:50one of two options are going to happen.

1:52Either we're going to connect these to the network core,

1:55or we're going to connect them to what we call

1:57an aggregation layer switch.

1:59So the aggregation layer switches, really deployed

2:01in pairs, would be two usually higher-performing,

2:05higher-featured switches, and they're

2:08going to connect all of the access switches.

2:10So as we draw the cell, let's say we have even more access

2:13switches down here.

2:14All of these switches are going to have connections up

2:17to these two aggregation switches.

2:20Now, if you're studying or already are aware

2:22of the three-tier architecture-- which is core, distribution,

2:26and then access--

2:27you may start to think, wait a second,

2:29this is sounding familiar.

2:30And it should, because aggregation

2:32is really just another word for a distribution layer.

2:35The only difference is that we call it aggregation so

2:38that we know, if somebody calls it aggregation,

2:40we're talking about the data center distribution layer.

2:43Now, these two switches are connected to each other,

2:45and then they extend up to the network core.

2:48Now, you might be thinking, well, Jeff,

2:50you said I can detach these straight to the core, right?

2:53And yes, that is absolutely an option.

2:54It is called a collapsed core design.

2:56In the collapsed core design, we actually

2:58don't have an aggregation layer.

3:00Instead, we bypass that by connecting our access switches

3:03straight up into the core.

3:05In fact, in small environments, we may take our servers

3:08and connect them directly to the core.

3:10There's a lot of different ways to build a network,

3:12and certainly, as long as these servers down here

3:15have network connectivity, then at the end of the day,

3:17we're providing the services that we need to provide.

3:19The reason that we have these more advanced architectures

3:22is to scale out for larger environments, because imagine

3:25trying to connect 500 servers up to a single network core layer.

3:29That probably wouldn't work too well.

3:31So we need to have these architectures in place

3:33in order to support those types of environments.

3:35So Let's create a clean drawing, here.

3:37We know that we have access switches at-- we'll put them

3:39at the bottom of our drawing.

3:41And we'll draw four out here.

3:42And again, the servers connect straight

3:44into this access layer.

3:45So this is the access layer.

3:49Then those have connectivity up to a pair of switches--

3:52usually not more than two switches, even at larger scale.

3:55And this would be called the aggregation layer,

3:58and these have connections to every single access switch.

4:00In fact, there should be redundant connections-- one

4:03to each aggregation switch-- from each axis switch.

4:05And then if we have a core layer, then that core layer--

4:08which usually comprises two switches as well,

4:10and they're to connected to each other--

4:12there will be a full mesh of connectivity

4:14between the aggregation and the core layers.

4:16Now, our users who are accessing the applications and data

4:19inside the data center--

4:20these users are usually hanging off

4:22of the core in some fashion.

4:24So they'll be coming into the core layer via other switches

4:26usually, and then the core layer allows them to go through

4:29the aggregation and down into the access layer,

4:31connecting into the servers and storage

4:33that they need to connect with.

4:35This network flow has a very specific name

4:37when we're talking about a data center,

4:39and that is called North-South.

4:41North-South because in our drawing here,

4:43it looks like it's going from north-- which is up here--

4:46down south.

4:48North-South traffic flow describes

4:49traffic that comes into the data center,

4:51and goes outside of the data center.

4:53Meanwhile, we have one other type of traffic

4:55we need to talk about, and that is called East-West.

4:58Now, East and West traffic--

4:59usually, on the compass, we're looking

5:00at left and right type of traffic,

5:02and that's what we're looking at here as well.

5:04East-West traffic describes traffic

5:05that stays inside the data center, often

5:07server-to-server communications.

5:09So maybe we have a server over here

5:11that's connected to a different access switch.

5:13Between the two, this is the heavier traffic flow.

5:16And so we need to worry, sometimes,

5:18about making sure we have enough bandwidth

5:20to support all the East-West traffic that we need.

5:22Now, speaking of bandwidth, we need

5:24a lot of bandwidth inside the data center-- to the point

5:26that servers usually connect, at a minimum, at 10 gig

5:29these days.

5:31We've actually improved that to 25 gig.

5:33So in a lot of data centers now, we have direct connection

5:36to our upstream switch at 25 gigabits per second.

5:40Now, the connections between access and aggregation--

5:43these tend to be 40 gig, but they can also be a 100 gig.

5:47Modern data center apps require a heavy amount of traffic,

5:50and so if we're pushing 25 gig per server,

5:53then we can very quickly realize that oh, yeah, 40 gig--

5:56that's a pretty heavy over-subscription.

5:58If I have two servers down here communicating at 25 gig,

6:02well, 25 plus 25 is 50.

6:04So yeah, 40's probably not enough.

6:07Now, the odds of both servers needing 25 gig at once

6:09is pretty low.

6:10However, that said, a lot of data centers

6:12are starting to build out 100 gig connections

6:15to the aggregation layer.

6:17And then between aggregation and core,

6:19that's really that North-South flow, and so a lot of times

6:21this is still built on 10 gig.

6:23But we clearly want to build this to support the users that

6:26are coming in, and so if 10 gig isn't enough,

6:28then we could migrate this to 40 gig, or even 100 if required.

6:32So you are now able to identify the primary components

6:34inside a data center network, where there's access switches,

6:37aggregation, core switches, and we even

6:39talked about different traffic flow types as well.

6:42I hope this has been informative for you,

6:43and I'd like to thank you for viewing.

0:01Anyone in the mood for pizza?

0:02[MUSIC PLAYING]

0:07By the end of the session, we will

0:09be able to describe servers and their role

0:11within the data center.

0:13I was out to dinner with my family the other night

0:15ordering some pizzas at a pizza place,

0:17and eventually somebody came out with our order.

0:20And that person is called a server.

0:23That person is called a server because--

0:25and when she came out, she said, here you go.

0:27It was serving up.

0:28She didn't literally say, here you go,

0:29but you know, hey, here's your pizza.

0:31And serving up that pizza to our family, because here you go.

0:35This is what you ordered.

0:36This is what you requested.

0:38Now, that's exactly why we call these big computers

0:40inside the data center servers, because users

0:43will want to access applications,

0:45or they'll want to access their data.

0:47And so when they want to access those things,

0:50the IP packets land on this computer, and the computer

0:53returns the necessary information.

0:55It serves it up.

0:56It says, here you go.

0:57This is what you've asked for.

0:58So let's go ahead and take a look

0:59at exactly how we build out a server, what it consists of,

1:02and then what it does, how it works within the data center.

1:05So I don't know about you, but I've

1:07built computers in the past for just home

1:09use, personal use and such.

1:11And I've got to say that computers these days--

1:13when you're talking about processors,

1:14we usually have one processor.

1:16And we might get four cores out of that one processor.

1:20And then, well, from a memory perspective,

1:22we might get up to 16 gigabits of memory--

1:25which is incredible.

1:26How powerful is that?

1:28That's amazing.

1:29And then from a hard drive perspective, I don't know.

1:32These days we could get a 512 gig flash SSD--

1:38could always go higher or lower on all of these specs,

1:40really, but this is probably pretty

1:42standard fare if you were to go out and buy

1:44a computer at a local retailer or something like that--

1:47order one online, however you get your computers.

1:49And of course, from a networking perspective,

1:51we might get 1 gig cards.

1:53You probably could get 10 gig these days.

1:56I don't know why you'd put 10 gig on your home PCs, I guess

2:00But hey, you know what?

2:02We could probably do it.

2:03So this is what we do for building out computers at home.

2:06Let's compare this to how powerful servers

2:08are within the data center.

2:10So in the data center, you almost always

2:12have two processors, but you could actually have four.

2:15So two to four processors, and here's the best part--

2:18I would say a standard server would

2:20come with 20 core processors.

2:22In fact, we can even go higher.

2:23I believe that Intel has 24 cores out there, possibly even

2:27pushing 32 at this point as of the time of this video being

2:30created.

2:31And so again, this is probably more

2:33of what you'd expect on average, but we can always

2:35push these specs higher as well.

2:38Now, for memory, we can get a wide range of memory.

2:40Just depends on what we need on the server.

2:42But I will say that these days, the minimum is probably

2:45about 384 gig on a single server,

2:49but we can go all the way up to 1 and 1/2 to 2 terabytes,

2:53just depending on how powerful we need.

2:55So we're talking about going from 16,

2:58all the way up to roughly 1,500 or 2,000 gig,

3:02which is absolutely incredible.

3:04Now, the interesting thing about the hard drives

3:06is that we might not actually have

3:08hard drives inside of a server.

3:10We're going to talk, later in the series,

3:11about storage arrays, but storage arrays

3:13are really where we put all of our storage.

3:15So we might have a small boot drive here,

3:18and that boot drive might be, I don't know, somewhere

3:21in the 64 to 128 gig range.

3:23But for the most part, we don't really

3:25put hard drives inside of servers anymore-- at least

3:28not unless the application requires it--

3:29which, some applications do.

3:31And last but not least, we already

3:32mentioned this in the previous video,

3:34but certainly 10 gig at a minimum,

3:36but we can even push 25 gig to servers.

3:39So sitting here looking at the two of these,

3:41servers are really, really powerful

3:43compared to what we're used to running here

3:45on our own home PCs, and laptops, and things like that.

3:49So why do these have to be so powerful?

3:50What value is there?

3:52Well, for the most part, it's because we're

3:54going to run a lot of different applications

3:57on a single server, and those applications

4:00are being served out to a lot of different users.

4:03And so the utilization on these servers

4:05is really going to be very, very high

4:08compared to what I do at home.

4:10And what about form factor?

4:11What do these servers actually look like?

4:13Servers used to take the standard desktop form--

4:15so we'd have what we call a tower, which is higher

4:18than it is wide.

4:19Pretty deep, too, for the most part.

4:21And these things would sit usually on the floor,

4:23on a shelf, or something like that.

4:25But they don't really stack very well.

4:27If you needed 20 of these things,

4:28how are you going to do that?

4:30Maybe turn them on their side and stack them up or something.

4:32Fortunately servers took on a different form factor

4:34a number of years ago, and that is

4:36to make them fit inside of racks.

4:38And so we get what we call pizza box servers.

4:40And the reason they're called pizza box servers

4:42is because they're really long, and just a little bit tall,

4:45just like a pizza box.

4:47In fact, they stretch back really far as well--

4:49even further than the tower servers

4:51did, because we don't have as much real estate vertically,

4:54so we have to stretch everything out horizontally and just

4:57flatten them out.

4:58So a server like this, that I've drawn right here-- this

5:00would be what we call a single rack unit.

5:02A rack unit is 1.75 inches in height.

5:06Now, the reason we call it rack unit

5:07is so that when we're building a rack--

5:09over here on the right.

5:10I'll draw this out.

5:11We might have a 42 rack unit rack,

5:15and now we don't to worry about inches and centimeters

5:17and conversions and all that.

5:19We can simply say, you know what?

5:20There's 42 slots up and down on this,

5:24and so if I have a 1-rack unit server,

5:25I know that that could get installed

5:27into the rack between two of those markers.

5:30Now, there are larger servers for sure.

5:32We can have 2-rack unit servers, so it'll be even taller.

5:35And hey, you know what?

5:36We can even have 4-rack unit servers,

5:39and there are devices within the data center

5:40that go up to 8-rack units and beyond.

5:43The reason we make these larger is because of real estate.

5:45If we're going to actually install 1 and 1/2 terabytes

5:48into our server, that's a lot of memory-- what we call DIMMs.

5:51These chips that get inserted into the motherboard--

5:55we have to plug them in.

5:56And so with 2 terabytes of memory, we have a lot of those,

5:59and so we need a lot of real estate.

6:01So as always, it just depends on what applications

6:03we're deploying onto the server, and how powerful

6:05it needs to be.

6:07And with that, you can now describe servers and the role

6:09within the data center.

6:11I hope this has been informative for you,

6:12and I'd like to thank you for viewing.

0:00The topic for today is "hype visors."

0:02[MUSIC PLAYING]

0:08When we're done here, we will be able to describe the function

0:11and benefits of a hypervisor.

0:13So calling this a "hype visor" is probably a stretch.

0:16Number 1, it's not a visor.

0:17It's a hat.

0:18And number 2, maybe it's not that hype.

0:20I don't know.

0:21But either way, we're going to be talking about hypervisors.

0:23And the word hypervisor itself comes from supervisor-- which,

0:26supervisor means somebody who looks over others

0:29who you're supervising, while hyper

0:31is one grade higher than super.

0:34And so a hypervisor is sort of a supervisor of supervisors.

0:38Well, hypervisors allow us to deploy multiple operating

0:41systems to a single server.

0:42So let's dive into the whiteboard

0:44and see how exactly this works.

0:46When we're on our home PCs, we have an operating system.

0:49A lot of times this is Windows.

0:50Sometimes it might be macOS, or Linux,

0:52or whatever flavor of operating system we're running.

0:55And we like to deploy a lot of applications

0:58onto this operating system.

1:00So for example, right now I have my screen recording tool going.

1:04I have my video editing software going.

1:06I've got my web browsers up.

1:07I've got so many different applications

1:09going all at the same time, all running on a single operating

1:12system.

1:13And that's also how it used to work in the enterprise space

1:15as well.

1:16For companies, we deploy a single operating system

1:18and deploy as many apps on top of that operating

1:21system as possible.

1:22The problem with that is that businesses

1:23tend to have more stringent requirements

1:25on their application's responsiveness,

1:27and how it's running and such.

1:29And so application vendors got tired

1:31of people calling in with issues when

1:32they had multiple apps running on the same operating system,

1:35and these are in contention with one another for resources.

1:39So instead what we decided was that OK,

1:41every single app requires its own operating system instance.

1:45What that means is if you have 30 different applications

1:48in your environment, then you're going

1:50to need 30 different operating systems--

1:52which, by extension, really--

1:53all this is running on hardware, right?

1:55So that means 30 different servers

1:57running your environment.

1:58Well, as you can imagine, it's not

1:59ideal to run 30 different pieces of hardware

2:02for 30 different applications-- as much as anything,

2:04because this one application might run 10% of the CPU,

2:09and maybe 40% of the memory, or something like that.

2:13And then the next server might be the exact same--

2:1510 CPU, and 40, 50 memory.

2:17I don't know.

2:18But either way, we're not fully utilizing

2:20any of these servers with this situation.

2:22It's why we run multiple apps on a single operating system,

2:25because we have enough resources to run those apps.

2:28So the question is, how do we get multiple apps back

2:31onto a single piece of hardware?

2:33If we require a single app to be mapped to a single operating

2:36system instance, that can be pretty tricky.

2:39How do we get, for example, multiple copies

2:41of Windows installed onto a single piece of hardware,

2:43on a single server?

2:44Well, that is where the hypervisor comes into play.

2:47A hypervisor-- think of it as an operating system,

2:50but it's sort of an operating system of operating systems--

2:53which, again, as I said in the intro,

2:55a supervisor of supervisors.

2:57That's why we call it a hypervisor.

2:58Oops, no need for an S there.

3:00And so this hypervisor can be installed onto a single server,

3:04and that hypervisor will allow us

3:06to install multiple operating system instances.

3:09So this could be Windows, or Linux, or, again--

3:12well, probably not a whole lot of macOS

3:14going on in the data center.

3:15Primarily Windows and Linux, really.

3:17So now we have multiple instances of those operating

3:19systems, we can deploy as many apps

3:21as we can onto this server, while still

3:24meeting the one app per operating system instance

3:27requirement.

3:28And so now, in this case, as drawn out here,

3:30we have three different apps running

3:32on three different operating system

3:33instances on a single hypervisor on top

3:36of a single piece of hardware.

3:37So ideally, at this point, our CPU utilization

3:39maybe is more balanced.

3:40Maybe we're at 40% CPU, and maybe our memory

3:44is pushing 70% to 80%, something like that.

3:46We want to get our money's worth out of it.

3:48Obviously, we need to have enough overhead

3:50to allow for spikes.

3:51If we have a CPU spike, we want to be able to absorb that.

3:54But at the same time, we don't want be sitting at 10%.

3:56That's just kind of, again, a waste of money.

3:58We bought these awesome CPUs, and they're

4:00sitting there bored.

4:02Now, this process is called virtualization.

4:04And this stack right here is what we call

4:05a Virtual Machine, or a VM.

4:08This can also be called a virtual server,

4:09because it is a full operating system instance.

4:11It's going to have a server name and an IP address,

4:14and everything that goes along with a server instance.

4:16Now, there are many benefits to running virtual machines.

4:191 we already talked about, which is resourcing.

4:21But number 2, we can actually migrate virtual machines

4:24from one host to another.

4:25So a physical piece of hardware running a hypervisor

4:28is called a host.

4:30And so this is distinguished between a server,

4:32because now a server could mean a physical server,

4:34and it could mean-- if we're running a virtual machine up

4:37here, it could mean a virtual server.

4:39So we can't really used the word server anymore,

4:41so we use host and VMs, which can be called guests.

4:44So let's look at a scenario where the host on the left

4:46is running out of resources.

4:48Maybe its memory is really high-- like 90%, or something

4:51like that.

4:52And this host over here is sitting there

4:54with memory at 50%.

4:56And so what should we do?

4:58Well, it'd be nice to load balance this.

5:00And so let's migrate a virtual machine across here.

5:02It's called Live Migration.

5:04And we put the virtual machine, now,

5:06over onto the other host instead of the first one.

5:09This is also really beneficial if the host on the left

5:11goes down for any reason, and now we

5:13need to make it a resilience play.

5:15Let's get those virtual machines back online.

5:17And if we think about back in the old days,

5:18with physical hardware-- what we call bare metal hosts, where

5:21one application's on one operating system,

5:23is on one piece of hardware-- if a server goes down,

5:26we've just lost access to that operating system.

5:28We've lost access to that app.

5:30And getting that application back online

5:32is going to be really tricky.

5:33Well, in a virtual environment, this is what we do,

5:35and we can get back up-and-running really,

5:37really quickly.

5:38Now, there are many players in the virtualization space.

5:40The one that most people think about

5:42is VMware, because they were really the first ones

5:44to bring this to market.

5:45But then we also have Microsoft.

5:47Microsoft makes a product called Hyper-V.

5:50And so a lot of data centers now have

5:52started running their virtualization platform

5:54on Microsoft's Hyper-V instead of VMware.

5:56Citrix has a product called XenServer,

5:59and so that's certainly an option as well.

6:02And then last but not least, we have the open source version

6:05of all of this, which is called OpenStack.

6:07And OpenStack brings a lot of different services,

6:09including a hypervisor.

6:10But regardless of which vendor you deploy into your data

6:13center, you're going to find that they

6:14share a lot of the same features and benefits.

6:16Really, everything we've talked about today

6:18is available with any of these vendors.

6:20So there you have it.

6:21You're now able to describe the function

6:23and benefits of a hypervisor.

6:24I hope this has been informative for you,

6:26and I'd like to thank you for viewing.

0:00Let's talk about storage arrays, or is it SANs?

0:03[MUSIC PLAYING]

0:09By the time we're done here, we'll

0:10be able to describe the features and benefits

0:12of network-based storage arrays.

0:14Storage arrays are often inaccurately

0:16referred to as SANs.

0:18So if you've ever heard anybody say,

0:19hey, the SAN is down, well, what they are probably

0:22saying is that the storage array itself

0:24is having some kind of issue.

0:26The Storage Area Network, or SAN--

0:27that is how we access the storage array.

0:29So storage array's where the hard drives live,

0:32and Storage Area Network is how we access that array.

0:35So in this Nugget, we're going to take a look at the storage

0:38arrays themselves, and in the next Nugget,

0:40we'll be taking a look at the Storage Area Networks.

0:42We've already talked about how servers are just

0:44more powerful PCs.

0:46And of course, those servers then

0:47can have hard drives in them, usually denoted

0:49by this circular formation here that I just

0:51tried to sketch out.

0:52So if this is what we call a hard drive-- which would either

0:55be a hard disk drive representing a magnetic disk,

0:58or a solid state drive, which is the more modern way of storing

1:01data and faster electronic media,

1:03we can actually have a lot of these within a server.

1:06Cisco, for example, makes a server that

1:07has 24 hard drive slots in it.

1:10That's a lot of hard drive slots.

1:12Whereas with most PCs that we use at home in our laptops,

1:14for example, we really only have one hard drive.

1:16Now, we also mentioned that these days, servers

1:18are deployed without any hard drives inside of them.

1:20However, originally, this was the only way to build servers--

1:23was to put hard drives directly in there.

1:25We had a couple of problems arise.

1:26First of all, if this server were to get fried electrically

1:30or something dropped on it, or who knows what happens,

1:33we might actually lose the data.

1:35And losing the data is not ideal just because we lost a server.

1:38And the other issue is very similar to that concept

1:40of virtualization, where I might have, let's just say,

1:4310 terabytes of storage on this server,

1:45and I have a second server down here that's also 10 terabytes.

1:49Well, if the top server here is at 9.5 out of 10 terabytes

1:53and the server down here is at 1.5 out of 10 terabytes,

1:57well, holy smokes, that's a lot of unused storage here

2:01that would be lovely if I could actually use that up here.

2:05So these problems, in addition to some others,

2:07are why we went with this idea of taking the hard drives out

2:10of servers and putting them into what we call

2:13a network-based storage array.

2:16So we can have a network-based storage array here that has up

2:18to--

2:19I mean, name a number--

2:2050, 70, 100 hard drives.

2:22Just depends on the vendor and the model that we're deploying.

2:24We can support dozens, if not hundreds of hard drives,

2:27within a single storage array.

2:29This gives us a number of benefits.

2:30For one, it allows us to protect these disks better.

2:33If a server were to go down, then it's

2:35not taking any data with it.

2:36But at the same time, obviously we

2:38need to make sure that this device doesn't go down.

2:41And so we can build a lot of redundancy

2:42into this, like multiple power supplies, and whatever else

2:45we need to in order to protect these disks.

2:48We mentioned utilization.

2:49Utilization is so much better now,

2:51because we don't have to worry about that scenario

2:53where one server's completely underutilized, meanwhile,

2:56another server desperately needs storage space.

2:58So we're not going to waste any disks at all.

3:00We can utilize these disks to their absolute max.

3:03Finally, it allows us to add features as well.

3:05A lot of these storage arrays will

3:06come with backup and disaster recovery functionality.

3:09So really, just a lot of advanced

3:11features to help us wrap our arms around and protect

3:13and manage this data.

3:15By the way, these storage arrays are typically modular.

3:17So if we do start to run out of storage,

3:19we can simply add disks very seamlessly.

3:22A lot of times that comes in the form of a shelf.

3:24So we might have to add a shelf, which is another separate

3:26unit-- a chassis--

3:28and we can fill that with hard drives,

3:30and that gets attached to the base storage array.

3:32And so it just continues to expand on the storage that's

3:34supported on this platform.

3:36So one part we haven't addressed yet

3:38is simply that in a typical server or PC

3:41where we have hard drives installed,

3:42that hard drive has a dedicated connection to the motherboard--

3:46usually a connection that we call

3:48SCSI, which stands for the Small Computer Systems Interface.

3:51These messages are sent directly via the BUS to the motherboard,

3:55and so that communication is very high bandwidth and very

3:57high priority.

3:58So all of a sudden, if we're taking that hard drive out,

4:01we're throwing it out onto the network somehow,

4:04now we have communications this way.

4:06And say, OK, well, how do we make

4:08sure this stays high bandwidth and high priority?

4:11I mean, what does that network even look like?

4:14Well, that is the Storage Area Network, or SAN,

4:16and we're going to be covering that in the next video.

4:18And with that, you are now able to describe

4:20the features and benefits of network-based storage arrays.

4:23I hope this has been informative for you,

4:24and I'd like to thank you for viewing.

0:00We're having a sale on "ANs" today.

0:01We've got your LANs and your WANs

0:03and your MANs and your PANs, and also your SANs.

0:05[MUSIC PLAYING]

0:10By watching this video, we will be

0:12able to describe the characteristics of a Storage

0:15Area Network, or SAN.

0:16Our industry is inundated with TLAs, or Three Letter Acronyms.

0:20So we have Local Area Networks, which are LANs, and Wide Area

0:23Networks.

0:23Those are WANs.

0:24And in this video, we're going to be talking about Storage

0:27Area Networks, or SANs.

0:29So let's remember our scenario here.

0:31We have servers with hard drives in them

0:34that have direct connections to the motherboard.

0:36However, we want to pull those hard drives out of the servers,

0:39and put them into a storage array over here.

0:42The storage array's going to house

0:43all of the hard drives in our environment,

0:45so none of the servers actually need any of their hard drives

0:48anymore.

0:49However, what happens when we pull those hard drives out?

0:52We have to have communications between the servers

0:55and the storage array that, number 1, is lossless.

1:00What that means is we drop as few packets as possible,

1:02because within the server, this was lossless.

1:05We didn't lose any packets, so to speak.

1:08And second of all, we want as low latency as possible.

1:12This is a concern, because if we try to write to a disk

1:15and it takes a long time for that operation to happen,

1:17then that can cause problems with the applications that

1:20are trying to read and write that data.

1:22And as we can imagine, within the server,

1:24latency was no concern.

1:25We didn't have to worry about latency inside a single device.

1:29So how do we build a network that's

1:31as lossless as possible, that really provides as low latency

1:35as possible as well?

1:36Well, this is the challenge of a Storage Area Network,

1:39and this is why we have dedicated networks for storage

1:41in many cases.

1:43Now that we know the goals for SANs,

1:44it's important that we take a look at the protocols

1:46have been developed to help meet those goals.

1:49There are two primary types of Storage Area Networks-- one

1:51that uses TCP/IP, and the other one

1:54that uses a protocol that we call Fiber Channel, or FC.

1:56Now, IP addressing will usually use MAC address for ethernet

2:00at layer 2.

2:01So we have our layer 3 protocol here, and our layer 2 protocol

2:04here.

2:04Now, Fiber Channel and ethernet--

2:06they're like oil and water.

2:07We cannot combine them, for the most part, on a single switch.

2:11If we want to combine them onto a single switch,

2:13we have to make sure that we purchase the correct switch--

2:16one that's built for Fiber Channel and ethernet.

2:18Another option that we have is called fiber channel

2:20over ethernet, which will take an ethernet switch

2:23and run Fiber Channel across it.

2:25But even in this case, we don't actually use IP addresses.

2:28Now, as for IP address communication,

2:30there's one primary protocol that we typically will use,

2:33and that is a iSCSI.

2:34However, there are two other protocols

2:36that we should be aware of, which are CIFS and NFS.

2:40Now, it's outside of the scope of this conversation, really,

2:42to delve into the differences, but we

2:44can know that CIFS and NFS--

2:47this is called file-level storage,

2:49whereas iSCSI and Fiber Channel--

2:51these are what we call block-level.

2:54What this boils down to, essentially,

2:55is that if we have a server trying

2:57to connect to a hard drive using block level,

3:01then it's as if this hard drive exists inside the server.

3:05That server owns that hard drive.

3:07That's what it means to have block-level access.

3:10However, if we're accessing storage

3:12via a file-level protocol, then in that case,

3:15we don't have ownership of that hard disk.

3:17In fact, this is usually how we do file shares on the network--

3:20is allowing multiple machines to access this

3:23via these file-level protocols.

3:25So just to summarize here, for IP

3:27we have iSCSI, CIFS, and NFS.

3:31And for non-IP, we have Fiber Channel and FCoE.

3:38These are our primary storage protocols

3:40that we're going to use in a Storage Area Network.

3:42Now, when we build Storage Area Networks,

3:44a key consideration is what we call storage isolation.

3:48And what that means is isolating the storage network

3:50communications from normal, data-driven network

3:54communications.

3:55So I'm going to show you what I mean here.

3:57We have a server that's going to connect via its normal data

4:00ethernet NICs to the Local Area Network.

4:03And this would be how, oftentimes, our users

4:05are going to access whatever these servers are serving up--

4:08usually applications or data.

4:10Then if we want that server to have storage communication,

4:13then we'll typically connect via the Storage Area Network,

4:16and this is where it can communicate to the storage

4:19arrays, down here.

4:21Now, if this right here is what we call an air gap--

4:24meaning that we have switches up here and we have switches down

4:28here, so they're separate, isolated switches--

4:30well, now if there's a disturbance in the LAN--

4:33let's say we have a spanning tree

4:34failure, or some major issue.

4:36Well, even though the users can't access the servers,

4:39the servers did not lose access to the storage.

4:41That's a very good thing.

4:43And by the way, if we build the Storage Area Network

4:45with Fiber Channel or FCoE, this is typically

4:48how we have to build it--

4:49because again, Fiber Channel and ethernet,

4:51they're like oil and water.

4:52And unless we have very special switches,

4:54we can't actually mix those two.

4:56And so by purchasing Fiber Channel switches,

4:58we actually do get this storage isolation,

5:01which is very beneficial to us.

5:03However, now let's say we're using iSCSI,

5:05and we go to our bosses and we say,

5:07hey, we want dedicated switches to run iSCSI.

5:09So we want to make these-- even though they're

5:11ethernet switches for iSCSI, we're

5:14going to be running TCP/IP on them,

5:16we can still isolate them by deploying

5:18separate switches that connect to the storage arrays.

5:21But our boss comes back and says, hey, well,

5:23can't we just use these switches?

5:24We already own ethernet switches,

5:26so let's just use these.

5:27And that way we save a bunch of money.

5:29OK, well, I guess we could do that, right?

5:32And so we do that, and we connect our storage, now,

5:35up here.

5:36And so now our storage is hanging off the LAN

5:38in the same way our users are.

5:40Well, let's talk about that problem again.

5:41What happens if that LAN goes down?

5:44Well, the users have lost access,

5:45but now the servers have actually lost access

5:47to the storage as well.

5:49And if we think about that scenario

5:51where we have a server or a PC with a hard drive

5:53inside, I mean, imagine ripping this hard drive out.

5:57Servers and PCs do not do well when hard

5:59drives get ripped out of them mid-operation.

6:02And so this can cause a lot of corruption and a lot of issues.

6:05Oftentimes we'll have to pull from backups in order

6:07to recover from this scenario.

6:09So that's why it's always recommended,

6:10whether we go iSCSI or Fiber Channel or FCoE--

6:13it doesn't matter--

6:14have storage isolation.

6:16Have a separate Storage Area Network

6:18to make sure that it stays online

6:20in the event of a LAN failure.

6:21So now you are able to describe the characteristics

6:24of a Storage Area Network.

6:25I hope this has been informative for you,

6:27and I'd like to thank you for viewing.

0:00Does anyone else just love eating with a spork?

0:02[MUSIC PLAYING]

0:08By the end of our time together, we

0:10will be able to identify the architecture and value

0:12of a hyperconverged solution.

0:14For those who have never used a spork,

0:16it's taking a spoon and a fork and combining it

0:18into one utensil.

0:19It makes things like eating taco salads a lot easier--

0:23stab with the pointy end, and everything stays on the spoon.

0:26It's pretty great.

0:27So hyperconvergence is going to do something very similar.

0:30We're going to combine our storage

0:32and our server products into one.

0:34Now, if you're sitting there thinking,

0:36well, what's that going to look like?

0:37Wait a second, we're putting the storage back into the servers.

0:41Jeff, that's how we used to do things, right?

0:43And if you're thinking that, then yeah,

0:45you're absolutely right.

0:46But there are some key differences

0:48that we need to take a look at, and so let's go ahead

0:50and dive right in.

0:52Hyperconvergence is trying to clean up

0:53a little bit of a mess, because so far, we've

0:55talked about how we have servers, and then we have,

0:58also, storage arrays over here with all the hard drives

1:01in them.

1:02And then we have this network in the middle

1:04that we call the Storage Area Network, or the SAN.

1:06And this could be iSCSI, and it could be Fiber Channel.

1:09It could be FCoE.

1:11And this is mission critical.

1:12We have to keep this SAN online at all times,

1:14because the servers cannot lose access to their storage no

1:17matter what the circumstances.

1:19For some organizations, this is too complicated.

1:22We don't want to have to manage all of these moving pieces.

1:24then trying to keep them all online and configured

1:27correctly, and it just gets to be too much.

1:29So this is why see hyperconverged vendors

1:31showing up in the market and creating products

1:33that try to do away with the complexity,

1:35and make things simple again.

1:37So what we do is we take our servers,

1:39and yes, we put the hard drives back in them.

1:42So let's say we've got, I don't know,

1:4412 to 24 hard drives in each one of these servers.

1:47We have three servers here.

1:49Now, one of the problems of having hard

1:51drives directly in the servers--

1:53we remember that we can't actually use the storage

1:57from one server to another.

1:59So if this server here is at 90% capacity

2:02and we need storage disk, and this server down here is

2:05at 10%, that creates a problem.

2:08We can't use the storage that exists on another server.

2:11We also got all kinds of advanced features

2:13on the storage array.

2:14So we had things like backup, and disaster recovery

2:17features that we would want to take advantage of if we were

2:20to move to a different model.

2:22Well, fortunately, hyperconvergence

2:23gives us all of these features, and then some.

2:26The idea of hyperconvergence is we still

2:28have a software package that is running against the storage.

2:32So let's take a look at this storage array again.

2:34If we wanted to manage a storage array.

2:36We're going to log into a piece of software that

2:38is managing it.

2:39So this software-- call it a little bit of a wrapper.

2:42It's wrapping around these hard drives

2:43and managing them for us, doing things like RAID and, again,

2:47the disaster recovery, the backups and such.

2:50All of this is done not by the hardware,

2:53but by the software that is managing that hardware.

2:56So we're going to replicate this in a hyperconverged

2:58environment.

2:59We're going to have the same software

3:00package running right here.

3:02However, it's going to run against hardware

3:05that it doesn't actually own.

3:08These servers are going to be the hardware.

3:10They're going to be the storage array for us.

3:12So this software still gets wrapped

3:14around all of the hard drives in this environment--

3:16to the point, by the way, that if we

3:18were to log into these servers and take

3:20a look at either the operating system or the hypervisor,

3:22or whatever is installed on here,

3:23they will not see these hard drives,

3:26because it doesn't belong to them.

3:28This software is kind of a middleware-- middle firmware--

3:31that injects itself into the hardware

3:34and prevents the operating system

3:35from accessing the hard drives, because it wants

3:38to manage those hard drives.

3:40And since it's in the middle, now all of those servers have

3:43access to all of the storage, whether that storage exists

3:46on its own piece of hardware-- the server itself--

3:49or another server in the cluster.

3:51Now, one of the benefits of hyperconvergence

3:53is that it turns our data center into a little bit of a building

3:55block model.

3:56So let's say, again, in our example, we have three servers,

3:59and we're ready to expand.

4:01We want more storage, or more CPU--

4:03or maybe both, right?

4:05And so we say, you know what?

4:06We're just going to go out to our vendor,

4:08and we have 1, 2, 3 of what we'd call, maybe,

4:11appliances at this point--

4:12because it's servers and storage combined into one.

4:15And by the way, the software as well.

4:17So we don't need to worry too much about this.

4:19We simply go to them and say, hey, I need a fourth node.

4:22And so they ship me a fourth node.

4:23I plug it into the rack.

4:25I plug it up into the network, and it finds the other nodes.

4:28It adds the storage into the pool.

4:30It's a very seamless, again, building block-style

4:33of expansion for my hyperconverged environment.

4:36So again, think back to having servers and storage.

4:39So servers here, storage here.

4:41And if I need to do an expansion, well, hey, number 1,

4:45is it servers or storage?

4:47I might need to add another server onto here, which

4:49means that I have to do some complicated things in here,

4:53which is in the SAN.

4:54I might to rezone that SAN, for example.

4:56Plus I'll need to do some LUN masking over here

4:58on the storage array.

4:59And by the way, expanding a storage array

5:02is, unfortunately, usually not very seamless.

5:05So adding a shelf onto the storage array

5:07can be a little bit complicated, and can be a little bit

5:10expensive at times, too.

5:12And so the hyperconverged model can sit well

5:15with a lot of organizations for these reasons.

5:17Now, the downside to hyperconvergence can be that--

5:20in our example here, let's say we have four nodes,

5:22and now we're out of storage.

5:25We need more storage.

5:26And so I call my vendor up and say, hey, you know what?

5:28I just need more storage.

5:29How do I get more storage?

5:30They're going to say, oh, you know what?

5:32Let's just add a fifth node down here.

5:34Say, well, wait a second.

5:35I don't need another server.

5:36I'm perfectly good on compute and memory.

5:39I just need storage.

5:41But expanding a hyperconvergence can

5:43tie the servers and the storage together from a scaling

5:46perspective.

5:47Now, not all hyperconvergence products

5:48are the same in this regard, so it's just

5:50important to call your vendor and ask--

5:52especially if you're weighing multiple vendors-- to ask,

5:55in the future, how would I go about scaling these out

5:57independently if I had to do that?

6:00And with that, you can now identify

6:02the architecture and value of a hyperconverged solution.

6:04I hope this has been informative for you,

6:06and I'd like to thank you for viewing.

0:00Well, it was bound to happen eventually.

0:02[MUSIC PLAYING]

0:07By the end of our time here together,

0:09you'll be able to describe the purpose of backup and Disaster

0:11Recovery, or DR solutions.

0:13Our fictional organization has just

0:16lost access to all of its data in its data center.

0:19So something bad happened.

0:20Let's say maybe a ransomware attack hit,

0:23and we just lost access to all of our data.

0:25Or maybe our power is out in our data center,

0:29and we have no applications.

0:31We have no access to our data.

0:33We can't conduct business in this environment.

0:36So what do we do?

0:38Well, if we have good backup and disaster recovery practices,

0:41then we're in good shape.

0:42Backups and DR are often lumped together

0:44as the same solution, when in reality they're

0:46trying to solve two different problems.

0:48The first problem is how do we go back in time?

0:52In other words, we've just lost some data,

0:54and we need to go back in time to retrieve that.

0:56So hey, you know what?

0:57I've been there, where I've been the guy who's

0:59like, I think I accidentally deleted a file that

1:02was on the network share.

1:04Hey, IT, can you get that file back for me, please?

1:07And they'll go back into the backups,

1:09and they'll pull that out of the past.

1:12Really.

1:12Like they say, hey, you deleted this, but here it is.

1:14Congratulations, we've been able to go back in time.

1:17The other problem that we're trying to solve

1:18is, how do we get back online in the event of a large outage?

1:23So this would be the scenario where

1:25I mentioned a data center maybe is offline in some fashion.

1:28Maybe the power is out, or the network is down,

1:30or something really bad happened, and we're just down.

1:33We're not trying to go back in time at this point.

1:35We're just trying to get back online.

1:37So the top problem would be what we use backups to solve.

1:42And the second problem can be solved with good disaster

1:44recovery practices.

1:45So let's first talk about backups.

1:47So the point of backups is to take what we

1:50call point-in-time snapshots.

1:53And all that means is that we're going

1:55to be looking at our data, and we're

1:57going to be taking a snapshot, almost

1:59like a picture, of what that data looks like,

2:02and then we're going to store that.

2:03So for example, I might take a snapshot

2:05on Monday of what that data looks like,

2:08and then I'll take another snapshot on Tuesday,

2:10and then I'll take another snapshot on Wednesday.

2:12Now, these are called point-in-time

2:14because maybe it's at specifically 6:00 PM

2:17every single night I take my backups.

2:20And so if I go to my IT, let's say on Thursday--

2:24here's Thursday, and now it's 4:00 PM on Thursday.

2:28And I say, hey, I think I deleted my file,

2:32and I think I deleted my file sometime on Wednesday.

2:36Can you help me out?

2:37Well, since most organizations close at 5:00 PM,

2:40it's likely I went home by 5:00, which meant I deleted it

2:43before 5:00 o'clock.

2:44So was it actually in this snapshot?

2:48Well, probably not, which means that we

2:50have to go back all the way to Tuesday at 6:00

2:52PM in order to get that file back.

2:55Now, what if I put four hours of work into it Wednesday morning

2:57before deleting it?

2:59Well, my backups aren't going to help me,

3:00because the last snapshot I had was from Tuesday night.

3:03This is the idea and the concept of point-in-time snapshots.

3:07So you might ask, well, can't I take snapshots more often?

3:10And absolutely, you can.

3:11We could take snapshots every 15 minutes if we want to.

3:13And for some mission critical applications, that makes sense.

3:17However, the more snapshots that we take,

3:20the more our storage needs go up.

3:24And so we might be sitting on terabytes of backup storage

3:27if we take too many snapshots.

3:29So we have to find the balance between what

3:31the business needs, and how much it's going to cost to do that.

3:35So now you might say, OK, so my data center's down.

3:37I've lost access to my servers and everything like that.

3:39Can't I use my backups to get everything online?

3:42And backups can be part of a disaster recovery strategy,

3:45except consider this.

3:47Let's say I have my backups here in this appliance--

3:50whatever that is--

3:51and now I'm ready to put those backups back onto a server.

3:55So all my virtual machines are backed up here.

3:57This is great.

3:58All right, so where do I put these?

4:01I don't have any servers.

4:03I don't have any storage.

4:04I can't run.

4:05I can't have my users access the backups directly.

4:10They're highly compressed, and they just don't run.

4:12I mean, the backup appliance itself isn't powerful enough

4:15to serve those virtual machines up in most cases.

4:18So what do I do in this situation

4:20where I don't have those?

4:21And now we're talking about disaster recovery.

4:24So disaster recovery is all about, again,

4:26how do we get back online very, very quickly?

4:29And so we have various different levels of DR.

4:32So this might be a cold DR strategy,

4:34where we say, you know what?

4:35I don't have a way to get back online very quickly.

4:38I'm going to have to pull my backups.

4:40I might even have to go out and buy servers--

4:43maybe off of eBay, or wherever I can get them--

4:45pull them off a shelf if I happen to have some old spares.

4:49But I have to deploy my own servers before I can even

4:52restore the backup.

4:53So that's, like, a cold DR strategy.

4:56Then I might have a warm DR strategy.

4:58Maybe I say, hey, I do have a second--

5:00let's just call this a backup data center sitting over here--

5:03and I've got some servers running.

5:05And I don't really have all the software installed, everything

5:08like that, but hey, in the worst case scenario,

5:10I can take my backups, and I can deploy them

5:13onto those servers in my backup data center.

5:15Oops, backup data center-- that should be a BDC, not BDR.

5:19OK, so last but not least would be a hot DR strategy.

5:23Typically what this involves is having a primary data center,

5:26and that backup data center again,

5:29and we'll have servers running here, which is great,

5:31but we're also going to have servers actively running here.

5:35And especially in a virtualized environment, what that means is

5:38we're running a hypervisor on these servers over here.

5:42So not only in our production environment--

5:43of course we have the hypervisor over here as well, running--

5:46but also in our backup data center.

5:48And then we have what we call DR software that's

5:52running right here in the middle,

5:53and it's replicating my traffic, my data back and forth,

5:58almost on a live basis.

6:00What that effectively does-- it turns my backup data

6:02center into sort of a clone of my primary data center

6:06so that if my primary data center goes down

6:08for any reason, I can spin up very, very quickly

6:11in my backup data center.

6:13Now, the reality is is that this DR strategy here-- cold, warm,

6:16and hot--

6:17that is not something that I just choose from my entire data

6:19center.

6:20It's usually a per application basis.

6:22So I might have some applications--

6:24my low-tier applications-- in a cold DR strategy where

6:28my DR software here is not actively

6:30replicating those applications.

6:32That means that I have to have less hardware over here,

6:35and maybe even less licensing here.

6:36So that can save me a lot of money

6:38by putting my low-tier applications into a cold DR

6:40strategy.

6:42So in the event of a DR scenario, yeah,

6:43I could get my hot applications running.

6:46This would be my high tier, my mission critical applications.

6:49Those are going to be replicated,

6:51and they're going to be made live instantly-- more or less--

6:54in my backup data center.

6:55But I will still have to pull from my backups, which

6:57is a much slower process, to get the low-tier applications

7:00running in my backup data center.

7:02So while backups can be part of a DR strategy,

7:04for the most part, backups are best

7:06for helping us to recover lost data,

7:08whereas disaster recovery is all about getting us back

7:10online in the event of a data center failure.

7:13So now you are able to describe the purpose of backup

7:15and disaster recovery solutions, as well as to understand

7:18the difference between them.

7:20I hope this has been informative for you,

7:21and I'd like to thank you for viewing.

0:00When you were growing up, did your teacher ever

0:01tell you to get your head under the clouds?

0:03Well, guess what?

0:04It's time to get your head back up there.

0:07[MUSIC PLAYING]

0:12By the end of this video, we will

0:14be able to describe cloud computing and the value

0:16it brings to server architectures.

0:19It's really hard to read any kind

0:20of industry update about technology

0:22without seeing all of this talk about cloud--

0:24cloud computing, cloud this, cloud that.

0:26We even see commercials on TV about, hey,

0:29you know what, this is the cloud.

0:31This is what the cloud can do for you.

0:33And even those commercials, they don't

0:35do a great job of explaining to us what exactly the cloud is.

0:38So by the end of this video, we're

0:39going to understand what that cloud is, and again,

0:42what it can do for us.

0:44We just defined what a virtual machine looks

0:46like inside of a data center.

0:47Remember, we're running an operating system

0:49and an application on top of that.

0:52And that virtual machine we said lives

0:53on a physical piece of hardware called-- yeah,

0:56we could call it a server.

0:57But really, we like to call these hosts,

0:59because server can refer to the physical piece of hardware.

1:02But technically, this virtual machine is a virtual server.

1:04And so we don't want to be ambiguous.

1:06So any given host in our environment

1:08could run a lot of different virtual machines.

1:11And hey, guess what, if we have hundreds of virtual machines,

1:14we're going to need a lot of physical servers.

1:16Which causes us some pain points.

1:18For one, we have to manage all of these different hosts.

1:21For two, we have to go out and make

1:23what we call capital expenditures-- capex spends.

1:26That means we have to pay a bunch of money

1:28in order to buy these servers.

1:30This many servers can start to take up too much space

1:33in our data center.

1:34It can draw power.

1:35And power, by the way, is expensive.

1:36So we have to pay for that power.

1:38And in the end, wouldn't it be great

1:40if we had an option to not deploy our virtual machines

1:43to a physical piece of hardware inside of our own data centers,

1:46but to deploy a virtual machine inside of somebody else's data

1:50center?

1:51What I mean by that is, let's say

1:52that they deploy a bunch of servers,

1:54and they're the ones who have to buy the servers,

1:57and they're the ones who have to do

1:58the management for the servers.

2:00And they simply lease us the space on those servers.

2:04Well, this is what we call an Infrastructure

2:06as a Service or IaaS environment, which

2:09is one type of cloud.

2:10What this means is I can deploy my virtual machines

2:13onto their hardware.

2:14And so I might have virtual machines

2:16on a bunch of different servers in their environment.

2:18But to me, I don't care whether that's on one server

2:21or let's say three physical servers.

2:23It doesn't matter to me.

2:24What I care about is, is my virtual machine

2:26online and accessible?

2:28This changes the way I spend my money.

2:30Because now, just like a car lease,

2:31for example, I'm no longer purchasing hardware

2:34for my virtual machines.

2:35That would be the capex spend we talked about.

2:37But now I'm paying so much money a month

2:40based on my utilization of their equipment.

2:43And because I'm paying per month,

2:44this is what I call an operational expense,

2:46or an opex spend.

2:47Speak to our Chief Financial Officer, or CFO,

2:50about opex and capex, and boy, you'll get them talking.

2:53Because that's a really hot topic

2:55right now for a lot of organizations as to whether

2:57it's better to spend capex money or better to go with opex.

3:00Now, the interesting thing about these cloud environments

3:03is when they build a server, sure,

3:05my virtual machines might be on that server.

3:07Maybe I have two or even three virtual machines--

3:09how ever many virtual machines I spin up--

3:11happen to be on that one physical host.

3:13But I am not the only customer in this environment.

3:16So they may have other customers that

3:18are running virtual machines on the same piece of hardware.

3:20Now, part of what we call being a cloud service provider--

3:23which is what we call the company

3:24responsible for managing this environment-- it is up to them

3:27to deploy the security to make sure

3:29that my virtual machines can't talk to other customers'

3:31virtual machines.

3:32So when we're talking about the cloud,

3:34we're really just talking about somebody else's hardware

3:36running in somebody else's data center,

3:38but running my virtual machines.

3:40The concepts and the technology of virtual machines running

3:42on hosts is the same.

3:44So what are some of these cloud service providers?

3:46Well, some of the most popular ones

3:48are going to be some of the technology companies

3:50we typically hear about.

3:51Amazon, for example, has a product called

3:53Amazon Web Services, or AWS.

3:56Microsoft also has a product that they call Microsoft Azure.

3:59Google is also in this game with Google Cloud Platform.

4:02And the list of other vendors outside of these three

4:04is just enormous.

4:05There are tons and tons of cloud providers out there

4:08that provide these services.

4:10Now that we know how cloud services work,

4:11we need to understand the different types of cloud

4:13services available to us.

4:15And these are going to be in two different categories.

4:17The first would be what we call the cloud deployment models.

4:20And the second would be what we call the cloud service models.

4:23There are four types of cloud deployment models.

4:25We have what we call a private cloud, which

4:27is what we talked about earlier, when we store our own data

4:30on our own servers.

4:31So we put virtual machines on our own servers

4:33inside our own data center.

4:34It's ours.

4:35We have full control over it.

4:36Nobody else is on it.

4:38Then we have public cloud.

4:40Public cloud is exactly what we just

4:41described with Amazon and Microsoft and Google,

4:44where it's somebody else's data center,

4:46and we're connected to that data center,

4:48and our virtual machines live on their infrastructure.

4:51Then we have a hybrid cloud.

4:53Hybrid cloud means that we're doing both.

4:55We've got our private data center here.

4:57We have connectivity up to a public

4:59cloud-- somebody else's data center.

5:00And we have virtual machines down here in our data center,

5:03and we have virtual machines up in the cloud.

5:05Last but not least, we have community clouds.

5:07These are primarily used by governments and academia.

5:10But typically what it is is you say,

5:11hey, you know what, I happen to have

5:13an infrastructure with virtual machines living here.

5:16And I'm going to share these resources with somebody

5:19else in my community.

5:20Again, think universities.

5:22Think government entities, where they're sharing resources

5:24with one another.

5:25Now, as for the cloud service models,

5:27we started with infrastructure as a service.

5:29This would be running our own virtual machines

5:31in somebody else's cloud.

5:32But it might also be borrowing their storage, for example.

5:36They might lease us so much storage space.

5:38There are all kinds of components of infrastructure

5:40that we might lease from another company.

5:42On the other end of the spectrum,

5:43we have something called Software as a Service,

5:45or SaaS--

5:46S-A-A-S. This means that really what I'm doing is I'm just

5:50running an application in somebody else's cloud.

5:52For example, I use a budgeting app.

5:54I log into this website.

5:55I point my web browser to https colon slash

5:59slash and then the URL.

6:00And I'm doing a lot of my budget tracking on this website.

6:04Well, that website is an application that

6:06lives in somebody else's cloud.

6:08So that is software as a service.

6:09I myself, I don't have to install that software

6:11onto my local computer.

6:12I simply log in with my web browser,

6:14and I'm accessing the application.

6:16Now, between infrastructure and software as a service,

6:18we have something called platform as a service,

6:20or P-A-A-S.

6:22Platform as a service is kind of in between,

6:24because with virtual machines, of course,

6:26I have access to the operating system,

6:28like we talked about in the last video.

6:29But with platform as a service, you

6:31know what, I don't necessarily need access

6:33to the operating system.

6:35But hey, you know what, I need more access

6:37than simply the application.

6:39This is very common with things like databases

6:41and other platforms, such as coding environments, that

6:43allow me to more directly access what I need for my job

6:47without having to worry about that operating system.

6:50So as you can see here, there's a lot to know about the cloud.

6:53Fortunately, here at CBT Nuggets, we have you covered.

6:55So be sure to go to that search bar.

6:57Simply type in cloud, if that's all you

6:59want to know more about.

7:00If you want know more specifically

7:01about the different platforms, than

7:03be sure to check those out as well.

7:04For example, Amazon Web Services and Microsoft Azure and Google

7:07Cloud Platform.

7:08We've got training on all of those different types

7:10of clouds.

7:11So be sure to check those out if you want more information.

7:14And with that, you are now able to describe

7:15cloud computing and the value it brings to server architectures.

7:18I hope this has been informative for you,

7:20and I'd like to thank you for viewing.

0:00HDDs, SSDs and NVMe.

0:02Oh, my.

0:04[MUSIC PLAYING]

0:09By the end of our time here together,

0:11we'll be able to describe HDD, SSD, and NVMe technology.

0:16There are a lot of different ways to store our data on a PC.

0:19For example, we have hard disk drives or HDDs,

0:21we have solid state drives or SSDs,

0:24and then we have the newest technology

0:25as of the time of this video at least

0:27called non-volatile memory express.

0:30However, that particular technology

0:32seems to fall into a different category than the first two.

0:35In this video, we're going to explore

0:37exactly what these different technologies do for us when

0:39storing our data.

0:41Computers need a place to store their data.

0:43And a methodology that we've had for a long time

0:46is using what we call hard disk drives.

0:48Hard disk drives are magnetic in nature,

0:50meaning that it stores its data using magnets.

0:53Now, all data in computers consists of 1s and 0s.

0:57That's it.

0:57We live in a binary world.

0:59And so we only need to be able to store

1:00these two types of information.

1:02So when it comes to magnetic hard drives,

1:04if we have a sector and we want to store a 1, for example,

1:07then we might charge this sector.

1:10So that's magnetically charged.

1:11And that tells the computer that, hey, this is a 1.

1:14Or hey, maybe we're going to leave it uncharged uncharged.

1:18[LAUGHS] And because it's uncharged, that means it's a 0.

1:21Hard disk drives consist of what we call a platter.

1:24This is a circular disk that has a magnetic material spun

1:27into this.

1:28The magnetic material is what we use

1:30to charge and uncharge based on the different sectors

1:33and the different values of 1s and 0s that we want to charge.

1:36Meanwhile, there's a head that reads the data,

1:38has to read this 1 or this 0.

1:40And the way it does that is it moves back and forth

1:42this way while the disk spins.

1:45If you've ever had the privilege of using a record player

1:47to listen to music, then you know

1:48what we're talking about here because the record spins,

1:51and the needle detects what the music is.

1:54The speed at which this disk spins

1:55is what we call the revolutions per minute, or RPMs.

1:58I remember when I bought my first computer,

2:00I got to choose between a 5,400 RPM and a 7,200 RPM hard drive.

2:05The faster the hard drive, the more

2:06performance my computer would have.

2:08Now, 7,200 is great from a commercial perspective.

2:10But in the enterprise space, when

2:12we deploy hard disk drives, we hope to get better performance

2:15than that.

2:15We call these 7,200s 7K drives because 7K stands for 7,000.

2:20So it's roughly 7,000 RPMs.

2:22But then we also have 10K drives and 15K drives.

2:26The fascinating thing is we have the technology

2:28to spin this faster than 15,000 RPMs.

2:31But if we spin it faster, believe it or not,

2:33it creates mini sonic booms around the edges of the platter

2:37because it is traveling faster than the speed of sound.

2:40So 15,000 is our speed limit, so to speak,

2:43as defined by physics.

2:44Now, solid state drives, they don't use magnetic medium.

2:47We use an electrical system to keep the data stored.

2:50The way we do this is we have a microchip.

2:52And within this chip, there are billions

2:54of electrical components that we call transistors.

2:57And what we do is we form cells out of these transistors.

3:00And we're able to charge this cell using electrons.

3:02And then we use special transistors to detect, hey,

3:05what's the charge of that cell?

3:07If that cell is below 50% charged,

3:09well, that would be a 0.

3:10And if that cell is over 50% charged, that would be a 1.

3:14So in both cases, we have ways of storing 1s and 0s,

3:17but the methodology by which we accomplish that

3:19is very, very different.

3:20Now, solid state drives have a lot of advantages

3:22over hard disk drives.

3:24And one of those is the size.

3:26This is a solid state drive.

3:28Hard disk drives are much larger because we

3:29have to have that magnetic platter in them versus, hey,

3:33this is just a bunch of microchips--

3:35what we call Application Specific Integrated Circuits.

3:38These ASICs are running the entire show.

3:41We don't have to have a big magnetic platter.

3:43And by the way, we don't have to spin anything.

3:45There's nothing mechanical about this.

3:47We don't have to have a motor that spins the disk.

3:49We don't have to have that head that

3:51moves back and forth reading the data.

3:53And so anytime we have moving parts in a system,

3:56those are the most likely parts to fail.

3:57And so the failure rate of solid state drives

3:59is much lower than hard disk drives.

4:02In addition to the size and the failure rates,

4:03the performance of solid state drives is way better.

4:06For example, we measure performance

4:07as something we call Inputs and Outputs Per Second or IOPS.

4:11That's how many times you can read and write something

4:13to a disk per second.

4:15Well, a 15K drive at the most, we

4:17could probably get about 200 IOPS out of a 15K drive.

4:20But a solid state drive at a minimum has about 2,000 IOPS.

4:24In fact, there are solid state drives

4:25that can go way higher than 2,000 IOPS.

4:273,000 5,000 10,000.

4:29But at a minimum, we're talking about a 10X increase

4:32in IOPS compared to hard disk drives.

4:34Now, this concept of performance is

4:35what's going to bring us to non-volatile memory

4:37express or NVMe.

4:39It's really not a new way of storing data.

4:42It's more of a way of accessing that data.

4:43This is why I said in the intro that it's a little bit

4:46different because whether we have

4:47a hard disk drive or a solid state drive,

4:51either way, our CPU--

4:52our central processing unit in our computer--

4:55needs to be able to talk to those hard drives.

4:58And the technology we've used to talk

4:59to those hard drives over the years

5:01have been technologies such as SCSI,

5:03which stands for Small Computer System Interface.

5:05Then we had SATA, which stands for Serial ATA Interfaces.

5:09And then we have SAS, Serial Attached SCSI.

5:12As we've progressed and improved our protocols,

5:14we've been able to increase the capacity for us reaching out

5:18to these hard drives, which has improved the performance.

5:20But when we migrated from hard disk drives

5:22to solid state drives, we did not change the protocol.

5:25For example, we still use SAS to connect to a lot of SSDs.

5:29Unfortunately, this SSD has way more

5:31performance potential than what SAS is capable of delivering.

5:35This connection right here has needed to be improved.

5:38And that is what we call Nonvolitale Memory Express.

5:42NVMe takes advantage of PCI Express or PCIe.

5:46This is the communications protocol

5:48that CPUs normally use to access different peripherals

5:51within a computer-- peripherals such as network

5:53cards and video cards and all of those things

5:55that require a lot of bandwidth.

5:57And so by leveraging PCIe instead of the protocols

6:00that we've had to rely on, this pipe

6:02now between the SSD and the CPU is much, much greater

6:06than it's ever been.

6:08This is a way of super charging our solid state drives

6:10to be actually able to deliver on the performance

6:13that they've been able to deliver for some time.

6:15We just haven't had the way of delivering that performance

6:18to the CPU.

6:18Now, one thing to note about NVMe

6:20is that if you're planning to install NVMe drives into, let's

6:23say, a server in your data center,

6:25you need to make sure that that server is equipped

6:27for NVMe because it does require special connections

6:31between the hard disk slots and the motherboard.

6:34And it has to be PCIe.

6:35It can't be any of the other technologies.

6:37So it's not just a matter of saying,

6:38oh, I'm going to upgrade my SSDs to NVMe drives.

6:42That's not the whole process.

6:43We have to actually make sure that our servers are

6:45equipped for it before we go out and procure those drives.

6:49And with that, you are now able to describe HDD, SSD,

6:52and NVMe technology.

6:53I hope this has been informative for you.

6:55And I'd like to thank you for viewing.

0:00Are you ready?

0:00Because you and I are going on a RAID.

0:02[MUSIC PLAYING]

0:08By the end of this session, we will

0:10be able to describe RAID technology.

0:12I don't know what we're going to RAID.

0:14Maybe my fridge?

0:15But when it comes to raiding our hard drives,

0:18we understand what exactly that means.

0:20And we don't get any favors from the industry,

0:22because we don't even all agree on what RAID stands for.

0:25It's either redundant array of, maybe it's inexpensive disks,

0:29or maybe it's independent drives.

0:31But fortunately, the technology is all really well established.

0:34In this video, we're going take a look at exactly what it

0:36means to RAID our disks.

0:39The idea of RAID is pretty simple on the surface.

0:41We're going to take a cluster of disks, which usually I

0:43draw like this, these little cylinders,

0:45but cluster them together for two purposes.

0:48One, to improve our ability to lose a disk.

0:51In other words, the redundancy of the system.

0:53And the other reason is to improve

0:55the performance of the system.

0:57We can do this in a number of ways.

0:59First of all, we have what we call RAID 0.

1:01And we're going to find out you're really quickly

1:03that we have a lot of different numbers.

1:04For example, RAID 1, RAID 5, and RAID 10.

1:07So we're going to talk about several of these in this video.

1:09Now, I'll go ahead and say right now that regardless of the RAID

1:12type, all of the disks inside of this cluster that we're

1:14building have to be identical.

1:16They have to be the same type of disk-- you know,

1:18hard drive or solid-state drive.

1:20And they also have to be the same size, the capacity,

1:22like 256 gig or 512 gig.

1:25The idea of RAID 0 is to get better performance out

1:27of our disks.

1:28And the way we accomplish that is

1:30by clustering two disks together and splitting

1:33the data between them.

1:34For example, if I need to store ABC, whatever that data is,

1:37I might store the A here, the B here, and the C back over here.

1:42What this allows is for my CPU to be

1:44able to pull from both of these disks at the same time.

1:47So in theory, I can pull that data ABC faster than I could

1:51if it was only on one disk.

1:52This is called striping data.

1:54We stripe the data back and forth-- again, A, then B, then

1:57C-- between the two disks.

1:59Next, we have what's called RAID 1.

2:01The focus of RAID 1 is not on performance,

2:03but its back on this concept of redundancy.

2:06And the way we do that is by mirroring the data.

2:09So instead of striping it, we are now mirroring it.

2:11So why do we call it mirroring?

2:12Well, we have two disks again in this example.

2:15And we're going to copy all of our from one to the other.

2:19So again, our example where I'm storing ABC on the left disk,

2:22well, hey, now I'm storing ABC on the right disk as well.

2:26That way when I'm pulling this data from the CPU,

2:28let's say that this hard drive on the left dies.

2:31Well, now I can get that data safely

2:33from the second hard drive.

2:34So RAID 0 and RAID 1 are completely

2:36different ways of raiding your hard drives together

2:38and they give very different results.

2:40One is, again, focused on performance.

2:42And the second is focused on redundancy.

2:45Notice by the way, we don't get any redundancy out of RAID 0.

2:48For example, the disk on the left dies.

2:50But we've just lost our blocks of data called Data A and Data

2:53C, we've lost that data.

2:55Now, the next range want to talk about our RAIDs 5 and 6.

2:58But before we do, we need to talk about this concept called

3:00parity.

3:01The idea of parity is that when I stripe my data-- let's say,

3:03again, I have A, B, C, and I'm striping that across multiple

3:06disks--

3:07then I perform an operation on the bits involved

3:10with A, B, and C, and I create a new set of bits

3:13called the parity.

3:14What the parity allows me to do is mathematically,

3:16even though it's complicated-- we're not going to drill

3:17into the details of this--

3:19but if I for example, lost access to whatever

3:21B is, that middle part of that data, that I

3:24can take the parts of the data that I do still have--

3:27in this case, A and C, and I add it to the parity,

3:30and I'm able to extract B back.

3:32So in other words, if we were to apply this concept to RAID 0,

3:35then in the event that I lost a part of my stripe data--

3:37let's say I lost B--

3:38I could actually get that back.

3:40So this is where RAID 5 and 6 come in.

3:42RAID 5 says I'm going to take a cluster

3:45of some number of disks.

3:46Let's just use the example I just gave here on the left.

3:49So let's say we have three disks.

3:50A, B, C is my data.

3:51It's going to get striped across.

3:53So that piece of data has been split up

3:54and stored on all three disks.

3:56And then I add a fourth disk.

3:58We're going to call that the parity drive.

4:00And this is going to store that mathematical operation of what

4:02that parity looks like what this allows

4:04is if I lose any one of these hard drives.

4:06So long as it's only one, I can piece that information

4:09back together by taking the data that I do have,

4:12using the parity drive, and rewriting that data.

4:16The concept of RAID 5 is that I have a single parity drive.

4:20And this is well and good, except I can only

4:21handle one disk loss.

4:23What happens if I lose a disk and before I go to replace

4:26that disk I lose another disk?

4:28Well, I've just lost all of the data in the entire system.

4:31And that's no good.

4:33As such, we have an option called Raid 6.

4:35RAID 6 says if we have the same scenario

4:37with three separate drives that were storing data on and then

4:40striping it across, that rather than having a single parity

4:43drive, We're actually going to have two parity drives.

4:45What this allows is for the failure of two disks.

4:48So if I lose any two disks in the entire system,

4:51whether it's a parity drive or a data drive,

4:53no matter what I'm able to extract

4:54the rest of the information, because I have two parity

4:57drives in the system.

4:58Now for both RAID 5 and RAID 6, it

5:00doesn't matter how many data drives I have.

5:02I could have 3.

5:02I could have 6.

5:03I could have 10.

5:04But the number of parity drive stays the same.

5:06I always have one parity with RAID 5.

5:08And I always have two parity drives with RAID 6.

5:10The biggest problem with RAID 6 is

5:12we take a major performance hit by deploying RAID 6.

5:15So even though it looks good from a redundancy perspective,

5:17RAID 5 is actually faster.

5:19Now speaking of performance, we do

5:20have one other option we need to talk about, which is RAID 10.

5:24RAID 10 takes the best parts of RAID 1

5:26and the best parts of RAID 0 and combines them together.

5:29That's why we call RAID 0 is striping the data.

5:32RAID 0 gives us no redundancy.

5:34In fact, we can use the word zero to help us remember that.

5:37RAID 0 provide zero redundancy.

5:39But RAID 1 gives us that redundancy with mirroring.

5:42But unfortunately, it doesn't give us the performance boost

5:44that we get from striping our data.

5:46So what happens when we combine these two concepts?

5:48So let's use our example where we have three disks.

5:51And let's go ahead and stripe that data

5:52across all three disks--

5:54A, B, and C. And again, with RAID 5 and 6 we

5:57would make parity drives over here and get

5:59our redundancy that way.

6:01But we're not going to do that with RAID 10.

6:03Instead what we're going to do is

6:04we're going to deploy the exact same number of disks.

6:07And this will be the mirror.

6:09And as such, the data is, again, striped A, B, and C

6:12across the right.

6:14So we're striping our data on both sides

6:16and we're mirroring between the two.

6:17Hey, guess what, that's the best of both worlds.

6:19We get a high level of performance.

6:22And we get a high level of redundancy.

6:24Because technically, we can lose a lot of different disks

6:26so long as it's not the same.

6:27For example, if I lose the disk that's B here,

6:30I could also lose an A disk and a C disk.

6:34And hey, I've got all of my data.

6:35So I could lose upwards of half my disks.

6:37However, the risk here is that if I lose the B drive here

6:40and the B drive over here, then I have indeed

6:42lost data at that point.

6:44So the redundancy is between 2 and many disks.

6:47So what's the downside to RAID 10?

6:48Well, the almighty dollar.

6:50Our ability to purchase this many hard drives

6:53might be limited by our budget.

6:55I know in the example we just gave

6:56we're talking about six disks.

6:58And we just looked at RAID 6 and it had five disks.

7:00I mean, that's not that much more.

7:02But what if this cluster right here consists

7:04of 10 disks or 12 disks?

7:06Well, now we have 12 disks here and 12 disks over here.

7:10Now, we have to have 24 disks in order to deploy RAID 10.

7:15Whereas with RAID 5 or RAID 6, we'd

7:16only need 13 or 14 disks, respectively.

7:19So as we can see, there are a lot

7:20of options when it comes to raiding our disks.

7:22But at the end of the day, it's all about

7:23increasing either the performance of our system

7:26or the redundancy.

7:27Or hey, in the case of RAID 10, we're increasing both.

7:31And with that, you are now able to describe RAID technology.

7:33I hope this has been informative for you.

7:35And I'd like to thank you for viewing.

0:00Well, we've reached the end of this skill.

0:01Let's go ahead and review some quiz questions here,

0:03and as we do, be sure to hit the Pause button

0:05and think through how you'd answer it

0:07before we review it together.

0:08First question here-- which isn't actually

0:10a question-- says sketch out a core aggregation access model

0:13and label the north/south and east/west traffic flows.

0:16So grab a scrap sheet of paper.

0:17Grab a pencil.

0:18Hit the Pause button, and go ahead

0:20and hit Play when you're ready to see how you did.

0:22OK, well, this comes back to the networking conversation

0:25that we had.

0:26Our servers down here are going to connect in at what

0:28we call the access layer.

0:30These access switches will sit usually

0:32on the tops of the data center racks,

0:34but they might be at the end of the row

0:36or the middle of the row, like we talked about.

0:38So these will connect up to what we call aggregation switches.

0:41The aggregation switches-- we usually

0:42don't have more than two of these in a data center,

0:45and they will connect to each other like this.

0:47And then the access switches will individually

0:49connect up to each aggregation switch in the environment.

0:52If you happen to draw more access switches-- which

0:54would be, hey, that's great.

0:55That's extra credit right there.

0:57These extra access switches would also have connections

1:00to each aggregation switch.

1:02Now, let me clear out this text here for a moment

1:04as we continue our drawing.

1:06So let's say, now, we have a pair of core switches.

1:08So these core switches-- which would really represent

1:11the rest of the network.

1:12Remember, our users live out here, and they're coming in--

1:15probably not directly attached to the core, but in some way,

1:17they're attached to switches that

1:19are then attached to the core.

1:21And so we need to connect those users into the data center,

1:23and the way we do that is usually

1:25by full mesh connections among the core switches

1:28and the aggregation switches.

1:30Now, these core switches are also usually connected

1:32with each other-- although we didn't really

1:33review that very much, where our focus has been

1:36on the data center network.

1:37So understanding that these servers--

1:38and also the storage, by the way,

1:40if the storage is IP and doesn't have its own dedicated storage

1:43area network--

1:44these would be connecting in via the access layer

1:46in order to get network access.

1:49OK, now we need to sketch out north and south

1:50flows, and east and west flows as well.

1:53So north/south would be between the users and the servers,

1:55or anything within the data center.

1:57This is what we call, again, north/south.

1:59East/west is traffic that stays inside the data center.

2:01So that would be a [INAUDIBLE] server.

2:03So if we have servers in here, and maybe servers over here,

2:06this would be an east/west traffic

2:08flow here between servers within the data center.

2:10So one thing to keep in mind in all this

2:12is that north/south flows are usually smaller

2:15than east/west flows.

2:16East/west flows are usually pretty big,

2:18and take a lot of bandwidth.

2:20So when it comes to securing these flows, when

2:23it comes to making sure we have enough bandwidth,

2:25the east/west can really give us challenges in that regard.

2:28OK, next question-- or, technically the first question,

2:30since the first one wasn't a question.

2:32What gets installed onto a server to allow virtual

2:35machines-- i.e. multiple operating systems--

2:37to get installed-- Linux, Windows, a hypervisor,

2:41or hyperconvergence?

2:44And if you chose C, you are correct.

2:47So remember that we have these stacks,

2:49where we have a one-to-one relationship

2:51between an application and an operating system instance.

2:54And we used to have to deploy that on a single piece

2:56of hardware so every server-- every physical server--

3:00could only really have one operating system instance,

3:02which meant it can only house one application.

3:04Fortunately, virtualization did away with that,

3:06and we can install this concept of a hypervisor.

3:10So this hypervisor here will get installed onto a single piece

3:13of hardware down here.

3:16And then from there, we can install multiples of these

3:18stacks-- these application one-to-one relationships--

3:21the application and the operating system.

3:23And each one of these stacks that we're drawing out here--

3:25really, this is what we call a Virtual Machine, or a VM.

3:29OK, next, what does a server serve up?

3:32Choose all that apply.

3:33So we have applications, data, networking, and pizza.

3:36I'm going to guarantee that pizza's not an option, so A, B,

3:40and C, pick and choose.

3:43And if you chose A and B, you are correct.

3:46Applications and data are the two most important parts

3:50of a data center-- really, what the data center houses.

3:53If you go to the business leaders, the owner or the CEO,

3:56what are they going to care about?

3:57Are they going to care about whether a particular switch is

4:00installed, or a particular server, vendor type is chosen?

4:03No.

4:03What they're going to care about are the applications and data.

4:06They're going to say, are my applications online?

4:08Do I have access to my data?

4:10So this is the function of a server.

4:12This is why we care about deploying servers

4:14into data centers.

4:16Next up, which system is designed

4:17to get applications back online as quickly as possible--

4:20backups, disaster recovery, hyperconvergence,

4:23and virtualization?

4:25And this one's a little bit tricky,

4:26because we have both backups and disaster recovery at play here.

4:30So we have to think through which one of these

4:32is trying to solve this particular problem?

4:34And the answer is B, disaster recovery solutions,

4:38because backups are really designed to go back in time

4:40and help us restore lost data.

4:42That's the purpose of backup systems.

4:45However, disaster recovery-- this focus

4:48is getting back online as quickly as possible.

4:50So if we're going to go get disaster recovery software

4:53packages, for example, that's what their focus is.

4:55It's not really to go back in time and restore data.

4:58It's just to get back online.

5:00And this is the last question.

5:01Which storage networking protocol cannot work

5:03on ethernet switches-- iSCSI, Fibre Channel,

5:06Fibre Channel over Ethernet, CIFS, or NFS?

5:11and if you picked Fibre Channel, you are dead on.

5:13Fibre Channel and ethernet-- we describe them as oil and water.

5:16They really can't run together.

5:17Traditionally, Fibre Channel switches

5:19will be separate pieces of hardware

5:21compared to ethernet switches.

5:22So if we have a server here that wants to connect up

5:25to the Local Area Network or the data side,

5:27that will be a LAN switch like we just described.

5:30This is really the access layer switch that we spoke about.

5:33And if we're going to deploy Fibre Channel,

5:35we have to have a dedicated Fibre Channel switch in order

5:38to make that happen.

5:39Now, as for the rest of the options,

5:41iSCSI happens over TCP/IP, which will be built on ethernet,

5:45so that's a guarantee here.

5:46CIFS and NFS are the same thing.

5:48They're built on TCP/IP as well.

5:50Fibre Channel over ethernet is a little bit interesting,

5:52because technically what the question's asking is,

5:55will it work on ethernet switches?

5:56And Fibre Channel over ethernet-- of course

5:57it's going to work on ethernet switches.

5:59However, it does require special ethernet switches.

6:02I can't just go out and get any ethernet switch and run FCoE.

6:06So we have to be very careful, when

6:07deploying Fibre Channel over ethernet,

6:08that we make sure we get the right switches.

6:10But at the end of the day, they will be ethernet switches,

6:13which makes native Fibre Channel--

6:14what we call the original Fibre Channel spec--

6:16that native Fibre Channel cannot run on ethernet switches.

6:20So how'd you do?

6:20If there's any content you need to review, be sure to go back

6:23and watch those individual videos.

6:24Otherwise, congratulations on completing the Identify Data

6:27Center Components skill.

6:28I hope this has been informative for you,

6:30and I'd like to thank you for viewing.