What's the Difference Between Cloud Metro vs. Traditional Metro?

00:04 Welcome to Juniper Cloud Metro AMA series, where we have

00:07 experts to answer your questions, any questions related

00:11 to next-gen Metro networks. I'm your host, Irene Zhang, Director

00:15 of Product Marketing at Juniper. Joining me today is Amit

00:18 Bhardwaj, Senior Director of Product Line Management at

00:21 Juniper. Amit, welcome. Today we have another question, what's

00:27 the difference between Cloud Metro versus traditional Metro?

00:31 So let's start with three big pieces, right? One, as you move

00:37 to the Cloud Metro architecture, there's going to be new

00:40 capacities required in the network, new capacity means new

00:43 speeds and feeds. You know, we're talking about 10GE, 25GE

00:47 or 100GE UNI, 400GE NNI. We're going to talk about integration

00:53 of the optical with ZR and ZR+ pluggablesr into the capacity,

00:58 into the routing platforms.

01:00 We're also going to look at new protocols in this scenario with

01:04 going towards EVPN and segment routing, as well as the driving

01:09 infrastructure from a protocol perspective, while also

01:12 supporting some of the traditional protocols. Because,

01:15 you know, we're not going just Greenfield. So there will be

01:18 some Brownfield to Greenfield scenarios.

01:20 And this is where, you know, for example, Junos stack comes into

01:23 play, you know, where we are able to support both the

01:25 traditional and the next-gen protocols into the JUNOS stack.

01:29 The timing is going to become a lot more critical in the

01:31 infrastructure because of the low latency use cases. So that's

01:35 the infrastructure piece.

01:36 Now when you start to look at one layer above, from

01:38 intelligence standpoint, you're looking at this network

01:41 architecture, where the traffic needs to be steered, wherever

01:45 the service is being delivered.

01:46 The service might be delivered at the cell site, one hop from

01:49 the cell site, or maybe two hops from the cell site. And these

01:52 workloads, you know, might move within the metro locations. So

01:57 the intelligence in the IP fabric, to have this capability

02:01 to steer the traffic wherever the services being delivered is

02:04 going to be critical.

02:05 Now, as you start to converge all the different use cases into

02:09 the common infrastructure.

02:12 That converge network requires a certain set of requirements in

02:16 the automation layer. Because underneath, you can use the same

02:19 protocols for delivering your broadband services, building

02:22 your 5G infrastructure, or you know, even doing things like,

02:27 separation of control plane user plane for certain applications,

02:31 driving the traffic to different workloads.

02:34 But as you drive all these use cases, now, how do we manage the

02:37 network infrastructure? And this is where network slicing comes

02:41 into the play. Because network slicing basically is a

02:44 technology that we use some of the VPN for the overlays, and

02:51 segment routing, or MPLS, for the underlay from a technology

02:54 standpoint.

02:55 On top of that we look at the quality of service attributes

02:59 deployed to guarantee some of the SLA s for this, along with

03:02 latency in the network. But the big piece of this is like how do

03:07 you orchestrate this?

03:09 That's why automation becomes key is that, orchestrating this

03:13 workflow for network slicing, where you can set up your

03:16 network slices, you can set up your overlays on the network

03:20 slices. And then you can assure each of these network slices in

03:24 an automated fashion. And then do a closed loop automation

03:27 around that. This is the automation infrastructure that's

03:30 going to be required to migrate towards the Cloud Metro

03:35 architecture.

03:36 So let me just summarize this thing again, right, new

03:39 capacities and architectures, you know, intelligence in the

03:43 network fabric, and the automation, you know, to drive

03:47 the all the use cases on a converged network, basically. So

03:52 that's just you know, how this is going to happen in the

03:57 service provider space.

04:01 What else do you want to learn about the next generation Metro

04:03 networks? Send us your question by email at

04:07 metroAMA@juniper.net. And join us at our monthly live AMA

04:11 session. You can find a date and registration link at the

04:16 description box below. See you next time.