Cisco Collaboration Design Essentials

0:00Today, companies have different options when it comes to their infrastructure.

0:03We have on-premises where the infrastructure is in data centers that the customer manages.

0:08We also have the cloud where we don't have data centers where infrastructure lives in.

0:12It's all in the cloud environment where someone else manages that infrastructure.

0:17And we also have hybrid, which is a little bit of both.

0:19So in this video, we'll talk about each deployment model

0:22and also why a company may choose one over another.

0:26Now, here we have the three deployment models or offerings

0:28when it comes to Cisco collaboration.

0:31Our first one being on-premises.

0:33With this deployment model, all the different collaboration services

0:36are hosted inside the customer's data center.

0:39So here the organization owns, operates, and maintains all the different infrastructure

0:45that provides all the different collaboration services to our end users.

0:52Now, our second deployment model is the cloud.

0:54With this option, Cisco is the one that's hosting and managing

0:58all the different collaboration services inside the WebEdge Cloud.

1:04And then you and I simply consume our user services,

1:09allowing managing users and also configuring different capabilities

1:13using the WebEdge Control Hub.

1:16Then we have hybrid where we're combining the on-premises and cloud model.

1:20Before you and I dive deeper into the different models,

1:23let's go over an analogy.

1:25You and I can think of a deployment model like choosing where you want to live.

1:30Having an on-premises system is like living in your own house.

1:35When you own your own home, you maintain everything.

1:38You have full control and everything sits on your property.

1:42A cloud system is like renting an apartment.

1:45Someone else handles the maintenance, the upgrades, and the building,

1:48you simply rent space inside the apartment.

1:52Then having a townhome is like having a hybrid system.

1:56So a townhome is where two homes are directly connected to each other.

2:01So here you own part of the building, not the entire building.

2:04Within a townhome, you're responsible for the items inside the home,

2:08the interior plumbing, the electrical,

2:11and you're responsible for your own interior repairs.

2:13In the United States, we have something called a A2A or homeowner association,

2:18and these folks are usually responsible for the roof, the outside walls,

2:22along with the landscaping.

2:25So a townhome really has some of the benefits of owning a home

2:28along with renting an apartment.

2:33So when someone says a deployment model or deployment offering,

2:36we're really talking about a model that defines

2:39where these collaboration services, when it comes to Cisco,

2:42are hosted along with who manages those services.

2:46So on-premises, these services are hosted in our data center.

2:51And here the organization owns, along with operates,

2:56and maintains the different applications and infrastructure

2:59within that data center.

3:02And this is where we would find applications like CUV,

3:05Cisco Unified Communications Manager.

3:07And CUCM is the primary call control platform

3:13that supports phone registration,

3:16along with call routing, dial plans, and other enterprise features.

3:20So when we say call control,

3:23we are referring to the system that manages how phone calls are set up,

3:26routed, maintained, and managed.

3:29So when we say call control,

3:31we are referring to the system that manages how phone calls are set up,

3:33routed, maintained, within our voice and video networks.

3:38In fact, let's take a road trip,

3:39and here we see the main page for CUCM.

3:44The application will be installed on a server

3:46somewhere in our data center.

3:48It's on top of a hypervisor,

3:50and the hypervisor will host the application.

3:53So CUCM, along with other applications you see here,

3:56are installed on hypervisors, on servers in our data center.

4:01So in this scenario, you and I are administrating and maintaining

4:05the applications, along with providing updates.

4:08You also have to maintain the server here,

4:11in our data center,

4:13along with power, backup, licensing, among other things.

4:17The takeaway being, you and I are responsible for the infrastructure.

4:22You also have a unique connection here,

4:24and that system is used for voicemail,

4:26along with unified messaging.

4:28You also have Cisco IAM and presence, not shown here.

4:34And this application provides presence information,

4:37along with instant messaging capabilities

4:39for Jabber and other clients.

4:41You also have Cisco Expressways.

4:43We have the core and the edge here.

4:46This allows devices outside the company network

4:50to communicate securely

4:51with internal Cisco collaboration applications or services

4:55without the need of a VPN.

4:59When users are outside the company network,

5:01so maybe in their home, possibly a hotel or mobile network,

5:06their devices normally sit behind firewalls and NAT,

5:08or Network Address Translation.

5:11Firewalls and NAT sometimes block communication

5:14to internal services in an organization like CUCM.

5:19Expressways here provide firewall and NAT traversal

5:22meaning that it safely allows that traffic through into our network.

5:27So what that means for us is that any devices outside the company,

5:31so it could be physical phones or Cisco Jabber or the WebEx app,

5:35can register to CUCM from anywhere.

5:41A user outside a company would be able to make and receive phone calls

5:45or access voicemail as if they were physically in the network.

5:48Again, all this without the need of a VPN tunnel.

5:52We also have Cisco Meeting Server,

5:54which is used for on-premises conferencing

5:56when it comes to video and audio meetings.

5:59Now, we also have connectivity to the PSTN or ITSP.

6:03So here we have a CUBE,

6:05and CUBE being an acronym for Cisco Unified Border Element.

6:09This is used for voice over IP,

6:12and this is used for voice over IP.

6:15This is how we get inbound and outbound calls

6:17from our internal network over to the outside world.

6:21We also could have a voice gateway,

6:25and a voice gateway connects traditional telephony networks

6:28to voice over IP networks.

6:31So this would be our T1 or E1,

6:34and this would be our T2,

6:36and this would be our T3,

6:38and this would be our T4,

6:40and this would be our T5.

6:42So this would be our T1 or E1,

6:45or PRIs, which is still common in some parts of the globe,

6:49and also FSO and FSX ports.

6:54So the voice gateway takes a traditional telephony

6:57and then connects it to our voice over IP networks.

7:02So a company would go with this deployment model

7:05in the event it needed full control of the network

7:08because of tight security boundaries,

7:11or maybe they have strict data regulations

7:13that they have to abide by.

7:15And because of that, they have to keep all their applications

7:18within their own data centers so they have full control.

7:25Alright, next we have the cloud deployment model.

7:28So here we don't have our own data centers,

7:30but Cisco will actually host and manage

7:33all the different collaboration services

7:35inside the WebEx cloud.

7:36In this model, customers like you and I

7:39and our end users consume the service,

7:41manage users, along with configuring

7:43different capabilities within the WebEx control hub.

7:46In fact, taking a road trip,

7:48we see the WebEx control hub right here.

7:52So we have the WebEx app.

7:56This is the application that's used

7:58for calling, meetings, and messaging.

8:01We also have WebEx calling.

8:03This is a call control platform

8:05that's provided and hosted by Cisco in the cloud.

8:08We have WebEx meetings.

8:12This is Cisco's cloud meeting platform.

8:16We have the WebEx control hub.

8:20This is where you and I would manage

8:22the administrator portal,

8:24so users, devices, policies, along with other settings.

8:27And we also have cloud PSTN options.

8:30So we have the ability to have PSTN connectivity

8:33via the WebEx cloud.

8:36So there are some reasons

8:38that an organization would choose cloud.

8:40One being we have minimal infrastructure to manage.

8:43When an organization has a data center,

8:46we have to buy, install, power,

8:48pool, and maintain physical servers

8:50with cloud services.

8:52The vendor, in this case, is Cisco.

8:54So we have to buy, install,

8:56power, pool, and maintain physical servers

8:57with cloud services.

8:59The vendor, in this case, Cisco,

9:01will handle the hardware for you and I.

9:03So we never have to buy any servers or rack them.

9:07In the event a hard drive goes bad,

9:09we're not running over to our data center

9:11attempting to swap over drives.

9:13We don't have to worry about space, power,

9:15pooling. You get the idea.

9:17With the cloud, we have rapid deployment

9:20because we're not ordering a server,

9:22waiting for it, have to rack and cable it,

9:24install the operating system and software.

9:25These services are already ready for us

9:27in the massive Cisco network

9:30that has data centers all over the globe.

9:32We also have lower operational overhead

9:35because you and I no longer

9:37are upgrading firmware,

9:39monitoring power consumption, for example,

9:41or managing hardware lifecycle management.

9:44For example, when a server

9:46goes end-of-life or end-of-support.

9:48We also have automatic updates and patching.

9:51Again, this is done by Cisco.

9:53They apply the security patches,

9:55system updates, feature improvements,

9:57along with optimizations.

9:59These are things that you and I would have to do

10:02when we have to manage our own data centers.

10:05A use case for going over to the cloud,

10:08let's say that our company

10:10has an aging phone system.

10:14Along with the aging phone system,

10:17we have different visual platforms for meetings,

10:19and we also don't have any unified messaging.

10:22Maybe the organization moves over

10:25completely to WebEx

10:27so they can have all these different services

10:29under one communication platform.

10:32Next, we have the hybrid deployment model.

10:35Now, here we're combining

10:37the on-premises call control, like CUCM,

10:40with cloud-based messaging and meetings.

10:43This gives the organization flexibility

10:45while moving over to cloud services

10:48at their own pace.

10:50For example, we can have the WebEx app

10:52register to CUCM

10:53for calling services

10:55while using the cloud

10:57for messaging and meetings.

10:59We could have WebEx register devices

11:01call internal extensions in our environment,

11:04or PSTN phone numbers

11:06using on-premises CUCM.

11:09But here the takeaway

11:11is that we're not strictly using on-premises

11:13or strictly using cloud.

11:15We call it hybrid

11:17because we're using a mixture of both.

11:19So where you could see this,

11:21let's say at this topology

11:23of headquarters,

11:25an HQ has to stand CUCM

11:27due to complex requirements

11:29that involve CUCM,

11:31maybe the call center,

11:33along with the PSTN.

11:35But maybe we also have 13 branch sites.

11:40So here we decide

11:42that we'll move these branch sites

11:44over to WebEx calling

11:46while keeping HQ on CUCM.

11:48So with hybrid,

11:50it gives us the ability

11:51to migrate over to the cloud.

11:55And we say gradually

11:57because A,

11:59we have some complex requirements at HQ.

12:01And I say gradually

12:03because we could have a technical requirement

12:05to keep CUCM

12:07along with other applications on-premises.

12:09Or maybe it's not a technical requirement

12:11or it's infrastructure in our data center.

12:13We have a contract for

12:15the last three years.

12:17So having this hybrid type of setup

12:19allows us to protect our investment,

12:21our data center,

12:23along with taking advantage of the cloud.

12:25So in this video,

12:27we went over the different deployments

12:29of Brains Francisco,

12:31either being on-premises,

12:33cloud, or hybrid.

12:35And now we know the differences between each

12:37along with what components

12:39belong to each deployment model

12:41and also why exactly an organization

12:43would choose one over the other.

0:00Sometimes we have friends over and we have dinner and we spend the night playing games.

0:03But when it comes to the planning portion of it, I have to make sure, or my wife to be honest,

0:08because she's a planner, that she has enough food for everyone who's coming.

0:12Because we don't want a situation where someone doesn't have food to eat or something to drink.

0:16That is like when you and I plan for our collaboration deployment, we want to ensure

0:21that we have enough resources in place to satisfy all the needs of our users.

0:25So in this video, we'll see how we can plan appropriately to ensure that our collaboration

0:29environment is able to provide services for every user on the network.

0:36So imagine, if you will, if you and I move into a new city

0:40and then build this main highway that has one street and two lanes going to the same direction.

0:46And maybe this leads to the next town over.

0:50Now at first, these two lanes are plenty for all the cars that travel over this road

0:56going over to the next town. So maybe at any given hour, there's probably a hundred cars

1:02that go over this road and we never get any traffic jams or congestion.

1:07But imagine if this town gains in popularity and now we have 10,000 users, or I should say people,

1:13that now live in this town. So now we have more people attempting to travel over to the next town

1:18attempting to travel on the same road with only two lanes.

1:22So now we definitely have traffic jams. So using my simple example here,

1:27this road was built without the vision of eventually 10,000 people will live in this town.

1:34So now we have people not too happy because they spent most of the day stuck in traffic

1:38waiting to get to our destination. But what if they built this road

1:43with the vision that years from now, many more people will live in this town.

1:48So instead of two lanes, they built a road that had five lanes. So the same 10,000 people

1:54have no problem with congestion and traffic jams traveling over to the next town. So here,

2:00the number of cars going over this road represents users and devices, in this case,

2:07endpoints like phones. And I should put in here. And the number of lanes

2:15represent server and gateway capacity.

2:21So we definitely don't want to run into a situation where business grows.

2:24And now we have more users on our network. But now our infrastructure that we have built

2:29cannot accommodate those number of new users. That's why planning is so important because

2:33if we build it right the first time, then we avoid headaches in the future. So headaches for

2:38you and I, because we're troubleshooting the issues that we're having. Also headaches and

2:43frustration for end users as they're trying to use our services.

2:50So we definitely have a few things to keep in mind when it comes to planning

2:54the capacity of our infrastructure. The first one being user count.

2:59So we're talking about the number of users or total number of users that need to access voice,

3:04video, messaging, and meeting. This will definitely impact the licensing requirement,

3:11hardware requirements, and also the virtual machine requirements.

3:16We also need to account for the endpoint count. So this will include the number of phones or maybe

3:23soft clients, along with video endpoints, along with the gateways, along with other peripheral

3:31devices that have to be registered. We also need to know the peak concurrent calls.

3:37And this means for us knowing the maximum simultaneous active calls during business

3:42peak hours. This is important because it will allow us to size our gateways or maybe our cubes,

3:50especially when it comes to calls going over to the PSTN or ATSP.

3:57So for example, if at 3 p.m. you average a peak of 500 calls,

4:04but the environment is only sized to take 100 calls onto the PSTN, then you have people that

4:10cannot place calls during their busy hour. In fact, when you talk to the ATSP, because you're

4:14trying to set up a SIP trunk, one of the questions they'll ask you is how many simultaneous calls

4:19do you need? Also, the number of calls is key to determining how much bandwidth do we need.

4:27Because the reality is that each active call on your network takes up a certain amount of bandwidth.

4:35And the number of calls also has an impact on licensing and hardware. You also have cluster

4:41node count. So using CUCM as an example here, we need to know the number of nodes that we'll need

4:49in order to determine load sharing but also redundancy. You also need to know the virtual

4:55machine resource requirement. So here we're talking about CPU, memory, along with disk usage.

5:04And we'll have to know this for each node in our infrastructure. And you'll definitely notice that

5:09the more users you have to support and the higher call volume, they require larger VM footprints.

5:16First, we'll focus on CUCM cluster sizing. Let's take a road trip to a document.

5:22This document here is called the Cisco Collaboration Sizing Guide for Release 15.

5:27And I'll include a shortcut here to this document below this video.

5:31And I want to scroll down to a certain section here.

5:35So right here, Cisco Unified Communication Manager. And let me scroll down some more. Let me give myself some more space.

5:42So under virtual nodes and cluster maximums, each CUCM cluster can support configuration and

5:48registration for a maximum of 50,000 secured or unsecured skinny or SIP endpoints.

5:55We also see if our cluster exceeds 1,250 phones, then at that point, two TFTP server nodes are required.

6:04Then we see the number of call processing subscribers,

6:08then we see the number of call processing subscribers. So these are nodes that actually

6:12take calls and the cluster cannot exceed four, plus four standby. So that's four active and

6:18four standby for a total of eight call processing subscriber nodes. Also note the total number of

6:24server nodes in a cluster, including the publisher. We have one publisher in a CUCM cluster,

6:30the publisher has a read-write version of the database. We have TFTP and media servers,

6:36may not exceed the number 21. So 21 is a maximum number of servers we can have in a cluster.

6:45So let me show you two redundancy models. We have one-to-one redundancy.

6:50So we have a CUCM publisher, the publisher node has the read and write version of the database.

6:57And this database contains information like the configuration, a lot of different endpoints and

7:02services, along with provisioning and system state information needed for call processing.

7:08So we have device information, user information, many of these we'll touch upon in a future video.

7:15We also have our TFTP servers, and these servers provide phone configuration files,

7:21along with firmware to the IP phones in the environment. Then we have our four call processing

7:28nodes, and I should say active nodes. These nodes register IP phones, they handle call setup

7:34and teardown, they apply a dial plan, they allocate media resources, and they also manage

7:38different call features, like transfer and conference. Again, all these features we'll

7:44look at in a future video. Remember a document that we're looking at said the number of call

7:50processing subscribers in a cluster cannot exceed four, and we have four right here.

7:55Now we also have four standby. So in the event we were doing maintenance here, maybe there was an

8:00issue with this server, any phones that were registered to that subscriber will fail over

8:06to the standby node. And here we have a one-to-one redundancy because each active server

8:11has one standby node. So let's just show you two-to-one redundancy. So here we still have

8:20active nodes, they're doing call processing, but each pair of call processing nodes has one

8:25standby server. So we see one here and one there. So here, in the event that one node is offline,

8:33for whatever reason, we have the ability to fail over to that standby node. But in the event this

8:39node goes down at the same time, and then we're in trouble here, because in this scenario,

8:45this one server or standby node cannot accommodate 25,000 endpoints. We have 12,500 here and 12,500

8:54there. If that's the case, then why do this? So it goes down to cost and also resource optimization.

9:05Because in a previous example, with one-to-one redundancy,

9:09each call processing active node had a standby node.

9:12You may say, well, that's the route I would take, and I would definitely agree with you if you had

9:16the following resources. Because we have to keep in mind, when it comes to having on-premises

9:22infrastructure, we have to maintain this. So we need VM resources, we need licensing, especially

9:29for the hypervisor, we need hardware, so servers to actually host our VM environment. Then we have

9:35our data center, so we have power, we have cooling, we need power, we need electricity,

9:40so we have power, we have cooling, we need CPU, RAM, and disk, we have to maintain these servers,

9:45and of course, we have to back them up. So with two-to-one redundancy here, this gives you almost

9:50the same protection, and I say almost as one-to-one, but we have the number of standby servers.

9:57So here we need fewer servers, it does provide lower ongoing costs,

10:02and it still provides strong redundancy.

10:04Now, Cisco will provide pre-recommended virtual machine configurations, or otherwise called an OVA.

10:12An OVA is a file format that contains a pre-built, pre-configured virtual machine.

10:18These OVAs are provided by Cisco, and the goal here is to simplify your deployment, because the

10:24OVAs are tested by Cisco, so therefore we have optimal compatibility and performance, ensuring

10:30that when you deploy a VM, they're matching Cisco's design, and using these OVAs helps us

10:37minimize issues caused by improper sizing or unsupported configurations. In fact, if you call

10:43tech support with an issue, and those guys and girls determine that you didn't use a proper OVA,

10:49at that point, you could be in an unsupported configuration,

10:52so therefore they can't do much for you until you properly configure that VM.

10:57In fact, going down this document a little bit more, to this table right here,

11:03we see the maximum number of endpoints supported per VM configuration.

11:07So we see small is 3,750, medium is 10,000, and then large here is 12,400.

11:17So in fact, let me take you to a different page here. So in this case, this is the OVA file that

11:22you would download when you install CUCM. So here we are in version 15, and if I scroll down a bit,

11:29you see small, medium, and large. And one thing that should stand out to you is as we go from

11:36small to large, then we see that the virtual CPU count gets higher, along with the speed of the

11:42CPU, the amount of memory, the virtual disk, and the virtual NIC is the same. So in this case,

11:50the virtual NIC is the same. In fact, on my desktop here, I have an OVA file that I used to build my lab, right here.

11:59Right-click, and open this up with Notepad++.

12:05All right, so here we see the word small. If I scroll down a bit, we see two virtual CPU,

12:11right here. The memory is 10 gig, then we have our disk of 110 gigabytes. If I scroll down a bit more,

12:19we see medium. We have two virtual CPU, 12 gigs of RAM, the same disk. Going down some more,

12:26we have large. This one has four virtual CPUs and 14 gigs of RAM, along with the same disk.

12:35In fact, let me log into my hypervisor.

12:39So here I am, logged into my ESXi host. I'm going to go to virtual machine.

12:44We'll create a new one. We have the option to deploy from OVA file. Let me hit next. I'm not

12:50actually going to build this, so I'll put demo there for now. If I click and select the OVA,

12:58it should be on my desktop, in the folder called files, and we see it right there. We hit open.

13:05If I hit next here, let's pick our datastore. So my host will read that OVA file and then offer me

13:14the option with small, medium, or large deployment. So once again, Cisco is ensuring that I'm

13:21deploying a virtual machine that meets their design and guidelines. So Cisco offers these OVAs to us.

13:29They are ready-made, tested, and fully configured to allow us to deploy these virtual machines

13:34to our enterprise environment. All right, so you and I have reached a good stopping point.

13:38I'll see you in the following video, where we talk about unique connections.

0:01All right, next we'll talk about UniConnection.

0:03UniConnection provides services like voicemail,

0:07along with auto-attendance.

0:13So here, a user can leave a voicemail for you and I,

0:16but they also have the ability

0:17to reach one of our menus within UniConnection

0:21and type option one for sales, two for support,

0:24and zero for the operator.

0:25So it can do call writing for us via the auto-attendance.

0:30So when it comes to UniConnection,

0:31we can have a cluster of two servers.

0:33So we have a primary and a secondary.

0:37In UniConnection version 15,

0:38we can have up to 20,000 users per server.

0:43Now when it comes to voicemail,

0:45UniConnection is active-active,

0:50meaning that both nodes here in the cluster

0:53answer phone calls, they take voicemail,

0:56they allow users to check their voicemail.

0:59So when it comes to call answering,

1:00message recording, and greetings,

1:02our cluster here is considered active-active.

1:05Now when it comes to that database for UniConnection,

1:12it's active-standby.

1:16Has a writable database

1:18that all administrative tasks are written to,

1:21and then that secondary node has a read-only copy.

1:24So we can see in the cluster,

1:26each node is capable of supporting 20,000 users.

1:30Unfortunately, this does not mean

1:31that we can support 40,000 users.

1:37For the reason being, UniConnection in the cluster

1:41is built for redundancy, not capacity.

1:47Both servers in the cluster

1:48maintain the same user database and the same message store.

1:52What's going on here is that we have replication

1:54between the two nodes.

1:56This cluster is redesigned for failover.

1:58So in the event that one node fails,

2:00the other node takes over call processing,

2:03along with message access,

2:05but the maximum supported users

2:07applies to the cluster as a whole, not per server.

2:11So all this to say that we can support up to 20,000 users

2:15in a UniConnection cluster.

2:17So we'll definitely dive into UniConnection later,

2:20but let me take you back to our document.

2:23So here we're still looking

2:24at our Cisco Collaboration Sizing Guide,

2:28and I'm going to scroll down to a portion of our document

2:33where they talk about UniConnection.

2:37All right, so here we are right here.

2:40We can see here that we can support

2:41from 1,000 to 20,000 users.

2:45So we have maximum number of ports.

2:47These ports are used to answer calls to UniConnection.

2:51We also have voice recognition sessions.

2:54So this will be used for speech-based interactions,

2:57like if I speak a name for UniConnection, the dial,

3:00or maybe voice commands in the menu.

3:03And then we have maximum test-to-speech sessions.

3:07Here, UniConnection may be reading a transcription

3:10or a system message to me.

3:13All right, if I scroll up just a little bit,

3:15you also have this table here.

3:19And here I want to focus on Cisco UniConnection

3:22and the Cisco BE 7,000.

3:25We see here maximum number of users supported

3:27in a digital networking solution, 100,000.

3:33A digital networking is connecting

3:35multiple Cisco UniConnection clusters together

3:38so they act like one big voicemail environment.

3:41This is done using either SMTP

3:44or Simple Mail Transfer Protocol

3:47to send messages between servers or VPN.

3:52That being voice profile for internet mail.

3:54Again, these two are used in digital networking

3:57to send voice messages to each other over a network.

4:01We also have maximum number of users

4:02supported in the HTTPS networking solution.

4:07So here we're using HTTPS APIs instead of SMTP or VPN.

4:14This is a more secure way of connecting

4:17UniConnection clusters together

4:19so that voicemail can be shared across sites.

4:23So I want to show you in the event you saw documentation

4:25with the number 100,000, where that came from.

4:29All right, since we're here,

4:30let's talk about Cisco Expressways,

4:33which I believe is earlier in this document.

4:36If I can find my mouse.

4:38All right, there we go.

4:40Let me scroll up and look for Expressways.

4:43All right, which is right here.

4:46All right, so we have small, medium, and large OVAs.

4:50And we can see based on the OVA

4:51how many registrations per cluster.

4:53So we have 15,000, 12,000, and 200.

4:58We also have the amount of video calls

5:00along with audio-only calls.

5:03As one would expect, if we went with a large OVA,

5:06we're throwing more resources at the virtual machine,

5:09therefore allowing for more registration,

5:11video, and audio calls.

5:13What we see here is CE-1300.

5:16This is a hardware appliance built specifically

5:19for Cisco Expressway-C and Cisco Expressway-E.

5:23Or in other words, a dedicated physical server platform.

5:26You see that we do have higher performance,

5:28so 30,000 registrations per cluster.

5:30We still have 2,000 video calls per cluster

5:33along with 4,000 audio calls.

5:37All right, so next I want to talk about a local gateway.

5:41So let me take you over to a document.

5:43So a local gateway is an on-premises device,

5:50most commonly a Cisco Cube,

5:52that acts as a bridge between a PSN or ITSP

5:56in your local infrastructure and whether it's calling.

6:04So a local gateway would sit in between that.

6:07A local gateway, also abbreviated LGW in documentation,

6:15allows organizations who are using WebEx calling

6:18to record calls to local carriers

6:20that are local to our infrastructure

6:23while still using cloud-based calling for users.

6:26In other words, this local gateway

6:28allows users to reach the PSTN

6:31by going through the local gateway

6:33and eventually to your infrastructure and now to PSTN.

6:37All right, so scrolling down a bit more,

6:39I want to show you a table here.

6:47All right, here we are, calling capacity requirements.

6:50Before we look at numbers,

6:51I want to talk about the difference

6:52between certificate-based and registration-based.

6:56Registration-based uses SIP registration

7:02from the local gateway over the WebEx.

7:04So it's simpler to configure

7:06and really designed for smaller brand sites.

7:09And when it comes to certificate-based,

7:11here we're using TLS authentication via signed certificates.

7:16And because of that,

7:17we have higher call volumes and user volumes.

7:21With registration-based,

7:22we see up to 250 concurrent calls

7:25and approximately around 2,500 user numbers.

7:29So we see this number jump up to 2,000 concurrent calls

7:33and also approximately 20,000 user numbers.

7:37Now here we see over-the-top and also interconnect.

7:40Now over-the-top means that the SIP trunk

7:42runs over the public internet, not through a private carrier.

7:47With interconnect, it's the opposite.

7:48We have a private direct connection to the carrier or cloud.

7:52And since we have the private connection,

7:54that's why we see these numbers jump up dramatically

7:56up to 6,500 concurrent calls and also 65,000 user numbers.

8:03All right, so I'll show you one more when it comes to Qube.

8:07Now I couldn't find a version 15 sheet,

8:09so I'll show you version 14.

8:12Now you and I will look at Qube later,

8:13but what Qube does is connect to your phone system,

8:16so CUCM, over to the PSTN in the cloud.

8:19So connecting one voice over IP network

8:22to another voice over IP network.

8:24Now Qube handles call routing,

8:26so it makes real-time decisions

8:28about where a call should go.

8:30It also provides security,

8:31so it's protecting the voice network

8:33with either TLS encryption or firewall or NAT traversal.

8:37And it also provides interoperability.

8:44So some examples are codec negotiation

8:46and also DTMF translation.

8:49So Qube will fix any incompatibilities

8:52so CUCM and other systems can communicate reliably.

8:57All right, let's scroll down to the bottom

8:59and we'll go into a table here.

9:08So we see with the Cisco 4000 series ISRs,

9:11and depending on the memory and the model,

9:14they will determine the maximum amount of sessions

9:17that we can support.

9:18We see 500 all the way up to 10,000.

9:22Along with how many new calls per second

9:25the Qube can reliably establish.

9:28If we scroll down a little bit more, we'll see here,

9:33the Catalyst 8300 Edge platforms

9:36are the replacement or successor to the ISR 4000 series.

9:40So these are the next generation of branch routers.

9:43In fact, at the time of this recording,

9:45the ISR 4000s are all end of sale.

9:48And the Catalyst 8300 is a supported replacement.

9:53All right, I also want to mention these here.

9:56We have the Cisco CSR 8000V,

10:00and CSR stands for Cloud Services Router.

10:03So these are both virtual routers

10:08that run as virtual machines

10:10in either public or private cloud environment.

10:13This is nice because it gives you flexibility,

10:15especially when deploying these in cloud environments,

10:18or maybe we're deploying a physical router,

10:20it's not practical.

10:23So in this video, we looked at different sizing options

10:25for unique connection,

10:27also Cisco Expressway,

10:29along with local gateways when connecting

10:31to WebEx into your local PSTN carrier,

10:35and also Qube, which is Cisco's unified border element.

0:00Great job on making it to the end of the skill.

0:02In this section you answered some questions and now you've come back to check with me.

0:05Let's get started and dive in.

0:08All right, for question one, match each deployment model with the best description.

0:13And our first one is hybrid.

0:16Hybrid really combines the on-premises CUCM call control

0:22along with a host of WebEx services like WebEx calling and WebEx meetings.

0:27We see that a lot during migrations from on-premises to the cloud.

0:32Or we have workflows that have to remain on-premises.

0:36All right, so let's see what our answers are.

0:40This one is good, a mix of CUCM-based calling and cloud-based collaboration features.

0:45So next we have cloud.

0:47These are services that are hosted within Cisco's WebEx environment.

0:51This model gives your organization quick setup,

0:54minimal hardware responsibility, and easy scalability.

0:59And then we have on-premises where we have different servers and applications

1:03hosted in our own data center.

1:06All right, we hit submit.

1:08Going to question number two.

1:11WebEx calling solutions utilizing a premises-based PSTN connection rely on which of the following?

1:18And the answer here is local gateway.

1:20A local gateway is an on-premises voice gateway

1:23that provides PSTN access for WebEx calling

1:26when a business keeps its phone lines or subchunks on-site.

1:30A local gateway acts as a bridge between WebEx calling in the cloud

1:34and the organization's local PSTN provider,

1:37allowing those endpoints that register to the WebEx environment

1:40to place and receive external phone calls through this local PSTN circuit.

1:45All right, so we choose the answer and hit submit.

1:48Going down to question number three.

1:50Which option best describes an OVA template used in Cisco collaboration deployments?

1:56An OVA template is a packaged virtual machine image

1:59that includes everything the application needs to run,

2:02including the recommended sizing for the RVM.

2:05Using the OVA template ensures that you and I,

2:09when we deploy our VM,

2:10have the exact specifications that Cisco supports.

2:14You will use this when installing applications like CCM or Unix.

2:21All right, so looking at our answers here, we'll go with this one.

2:27A portable archive that includes a fully prepared virtual machine

2:31with its operating system, application components,

2:34and preset configurations for streamlined deployment.

2:38All right, we hit submit.

2:39Let's go down to question number four.

2:43What does digital networking refer to in Cisco U&E Connection?

2:47Digital networking allows multiple U&E Connection clusters to communicate

2:52together so they act like a single distributed voicemail system.

2:56It is useful for large organizations that need several clusters put together

3:00for a unified experience when it comes to messaging and voicemail.

3:04All right, so looking at our possible answers, we'll go with this one.

3:07A method of linking multiple U&E Connection clusters so they function together

3:12as a unified voice messaging system with shared messaging capabilities.

3:16All right, we hit submit there, and we have our last question here.

3:21Which description best explains the purpose of the Cisco CE1300 appliance?

3:27The CE1300 is a physical appliance built specifically to host Cisco Expressway-C

3:33and Expressway-E.

3:35All right, with that information, we know our correct answer to be

3:38our first one right here.

3:40Dedicated hardware platform designed to run Cisco Expressway-C and E

3:44to support secure communication services inside and outside the network.

3:49All right, we hit submit, and we are good to go.

3:52All right, with that, this instruction is complete, and I will see you in the next skill.