Juniper Networks Service Provider Overview

00:00 (cool music)

00:06 - Hello everyone, and thank you for joining today.

00:08 My name is Sally Bament, and I am the Vice-President

00:11 of Marketing at Juniper Networks.

00:13 And today we're gonna have a discussion

00:15 about Juniper service provider strategy.

00:19 And joining me today is Raj Yavatkar,

00:21 Juniper's Chief Technology Officer.

00:24 Raj always a pleasure, and thank you for your time today.

00:27 I really appreciate it.

00:30 - Now, thank you Sally, looking forward to this.

00:32 - Absolutely. So, Raj,

00:34 I'm actually pretty excited about this opportunity

00:36 to share service provider strategy

00:39 with all those listening in.

00:41 Service provider customers around the globe

00:44 have been quite frankly,

00:46 the cornerstone and the foundation of Juniper's investment

00:49 and quite frankly, our success as well from day one.

00:53 And it is something we don't take for granted at all.

00:56 I know we have a lot to talk about today,

00:58 including doubling down on some new areas

01:01 that are going through transition like the Metro,

01:03 as well as some investments in new domains.

01:06 So maybe if we can get started and talk first of all,

01:09 about some of the market dynamics.

01:11 So a couple of questions for you to respond to.

01:16 First of all, experience of the end customer

01:19 is key in today's services economy,

01:23 and delivering a differentiated experience

01:25 is absolutely a priority

01:27 for the service providers we talked to.

01:30 But we hear it's also a challenge too.

01:33 Maybe if you can talk a little bit about why is that,

01:36 and what can they do to change that?

01:39 - So I think the way I look at is

01:41 the way the internet traffic is growing,

01:44 by one estimate, internet traffic is growing at 55% CAGR.

01:49 Which means in five years,

01:50 we'll have 10 times more traffic than we have today.

01:53 That's a tremendous growth,

01:55 and that's an opportunity in market.

01:57 Secondly, as people are moving workloads to cloud,

02:01 their expectations have changed from service providers.

02:04 And we are suggesting to our customer service providers,

02:07 different focus areas.

02:09 Let me list all of them here.

02:11 Let's start first with the scale.

02:13 As I said, the internet traffic is growing tremendously.

02:16 That creates an opportunity as well as challenge,

02:18 in terms of the scale at which

02:20 you have to deliver networking services and deploy networks.

02:24 And you of course need scale with elasticity,

02:27 because the cloud providers

02:28 have been providing elastic services,

02:30 which allows you to deploy infrastructure on demand.

02:34 Second focus area is automation and orchestration.

02:37 And for that, the service providers need to invest

02:40 in end-to-end automation,

02:42 orchestration of multi service networks.

02:44 Our third focusing suggest is integrated cloud.

02:47 You have to start by thinking of cloud native architectures.

02:50 Because you are delivering infrastructure.

02:53 That's going to be in terms of

02:57 either the cloud connectivity,

02:59 or the enterprise connectivity,

03:00 or you may have your own data centers in terms of

03:02 to be able to give cloud like services.

03:05 All of that requires you to think to

03:07 the cloud native architecture,

03:09 and all of the infrastructure

03:10 should be designed using integrated cloud mentality.

03:15 And finally security.

03:17 Security is even more important now than ever,

03:19 because the infrastructure you are deploying

03:21 and delivering to customers is very vulnerable.

03:24 So continue to focus on delivering secure infrastructure

03:27 is another focus area.

03:30 - It makes a huge amount of sense, Raj,

03:32 and I know we're gonna come back to each of those points

03:35 further along in the conversation.

03:37 And of course we hear a lot

03:39 about service provider revenues

03:41 being relatively flat or low growth rates.

03:45 But on a more optimistic front,

03:47 there are opportunities as well, right?

03:49 From your perspective,

03:51 how can operators monetize

03:54 their existing network investment?

03:56 Basically get more from what they have,

03:58 and also find new sources of revenue.

04:02 - So I think that multiple opportunities.

04:04 First of all, the space at which enterprises

04:09 are going through a digital transformation

04:11 creates a huge demand for enterprise services,

04:14 managed services that service providers

04:16 can only provide and monetize.

04:18 Another big development that has happened

04:20 is the growth of the 5G infrastructure.

04:23 5G infrastructure in like 2G, 3G, 4G,

04:26 which was all about mobile connectivity.

04:28 5G is first time bringing services oriented experience

04:33 to the network, by providing guaranteed low latency.

04:37 All that requires slicing support,

04:39 you have to be able to slice the network,

04:42 end-to-end basis, and assign slices

04:44 to different types of applications.

04:46 As you do that you offer differentiated quality

04:49 of service experience to each of those slices

04:51 that you can monetize.

04:53 - Yeah, and that is a great segue, Raj

04:56 into what we're doing here at Juniper.

04:58 And we have a very interesting vantage point,

05:02 because we have a very big seat at the table

05:05 with certainly service providers around the globe,

05:07 but also some of the largest cloud providers on the planet.

05:11 And of course,

05:12 enterprises too across a number of different verticals.

05:16 So maybe if you could walk through what Juniper is doing

05:20 to help service providers navigate through these challenges,

05:23 and also really capitalize on this digital transformation

05:27 that everyone is going through right now.

05:29 - Service providers are our most important customers.

05:32 We have been there with them

05:33 along the entire journey of the Juniper Networks.

05:37 So what we are investing in is really the three pillars.

05:41 We call them three pillars because those are

05:42 the three pillars that will really allow service providers

05:45 to achieve the goals and focus areas

05:47 that I mentioned earlier.

05:49 First is we are delivering scalable IP Fabric,

05:52 it's truly elastic, and scalable, and resilient,

05:55 and it can go from all the way

05:57 from the metro transport, to core networks and so on, right?

06:02 Secondly,

06:03 we are focusing on delivering cloud-first platforms.

06:06 By cloud-first what we mean is that

06:07 we think of every architecture has to be having

06:10 the cloud native characteristics,

06:12 with both virtualization,

06:13 as well as cloud native infrastructure.

06:16 Finally managed enterprise services.

06:18 We are combining the services like like SD-WAN,

06:20 Wi:Fi access, wired access, as well as security

06:24 to offer them in a managed offering,

06:26 in which you get guaranteed service assurance.

06:30 And those services can now be offered

06:32 by the service providers, to their enterprise customers

06:34 while managing and monetizing them.

06:36 And all these three pillars, we are really combining them,

06:40 by enabling some fundamental foundational technology,

06:43 such as security.

06:45 We don't think of security as an afterthought.

06:47 We build in security right from the start,

06:49 so that across

06:50 scalable IP fabric cloud first platform managed services,

06:54 security is built in to be threat aware,

06:57 to do malware detection,

06:59 as well as a classical firewall like technology.

07:02 The second enabler we are focusing on is automation.

07:05 We have invested in end-to-end automation

07:07 of all aspects of network lifecycle management.

07:10 So I think all those three pillars

07:13 with two enablers together

07:15 will allow our service provider customers

07:17 to focus on this focus area as I mentioned earlier.

07:21 - Raj, crystal clear.

07:23 So three solution pillows,

07:25 massively scalable IP services fabric,

07:27 cloud first platform approach,

07:29 and managed business services,

07:31 and threaded across all three of those,

07:33 connected security and intelligent automation,

07:36 crystal clear. - Thank you.

07:37 - No, it's great.

07:38 It would be great if we could sort of maybe do a bit

07:40 of a click down on each of them,

07:42 and let's take each one at a time,

07:44 and maybe if we can start with massively scalable IP fabric.

07:49 - So I think that massively scalable is important

07:52 because we are known for IP fabric.

07:55 But what we have done now is that we are enabling customers

07:58 to be able to build massively scalable fabric,

08:00 by providing whole range of the devices and systems,

08:04 across the span of 400G we are already leader

08:07 We are already leader, 800G platforms and so on.

08:09 While doing that, we're also making sure it's easier

08:12 to extend those, just like the data center approach.

08:14 We are providing similar ability for people

08:16 to be able to build massively scalable fabric,

08:19 by offering elasticity and being able

08:22 to automate to via automation

08:24 build capacity very easily, right?

08:27 'cause all of the traffic that comes from enterprises,

08:30 consumers is getting aggregated in Metro.

08:32 We also have convergence between wireline

08:35 and wireless networking coming together there.

08:38 So what we are doing for, we calling it a new Cloud Metro,

08:41 because what we're taking is principles

08:43 of the cloud-native infrastructure,

08:45 cloud native architecture.

08:47 And we are applying them

08:48 to both disegregation based on some cups,

08:52 where you separate out the control plane and the use plane,

08:54 but as well as aggregation of the traffic.

08:56 And with 5G, it's important to be able to use

08:59 the Metro as a place to aggregate all of the traffic,

09:02 while supporting these disegregated architectures.

09:05 So together this new Cloud Metro concept

09:07 is completely cloud native.

09:09 It's supports virtualization,

09:11 it's supports to disegregation.

09:12 And that will allow the customers to have lots of choices,

09:15 and to mix and match different types of infrastructure

09:19 to deliver cloud like metro impact.

09:23 - That's great Raj.

09:24 And obviously our second pillar

09:26 is a cloud first platform.

09:29 And I'm not sure if everyone fully understands

09:31 the scope of this.

09:32 So maybe if you could explain cloud first,

09:35 and more particularly

09:36 why it's so important to service providers?

09:39 - Cloud providers are investing

09:41 in this massively scalable fabric scalable fabric

09:42 we talked about our new Cloud Metro.

09:44 It's important for them to also be able to create elasticity

09:49 and capacity by using cloud-based architecture principles.

09:53 So by cloud-first what we mean

09:55 is that any platform we are trying to provide,

09:57 we think of it first as a cloud native architecture.

10:01 So for example, Telco cloud,

10:03 we are providing cloud native

10:06 Contrail based overlay networking,

10:09 Apstra based fabric management,

10:11 which comes with closed loop automation, right?

10:14 They're also doing the similar things

10:16 with respect to support for VNFs,

10:19 as well as containerized network functions,

10:21 and how they could be deployed very easily

10:23 in any environment using cloud-native architecture.

10:26 So that's what we mean by cloud-first.

10:27 - And then obviously the third solution area,

10:31 really is more related to how we're partnering

10:34 with our service provider customers as part

10:36 of their managed enterprise service offerings,

10:38 and quite frankly, helping their enterprise customers

10:41 through their own digital transformation journeys.

10:45 Tell us a little bit about our solutions in that area, Raj.

10:49 - So I think where we started as I said

10:51 with a Wi:Fi solution, which is already cloud managed,

10:55 and the way it works is that you connect every device

10:59 to the cloud to provide constant runtime telemetry,

11:03 that gives you constant operational insights

11:06 into every device, every system, every network.

11:09 Use that operational telemetry,

11:12 to feed into an machine learning engine,

11:14 to be able to derive insights

11:16 so that you can start predicting failures.

11:19 You can start detecting anomalies and failures.

11:21 But also more proactively, using predictive analytics.

11:25 You can also start taking remediating actions automatically.

11:28 So you really form a closed loop automation,

11:30 which is purely cloud managed,

11:32 not only for Wi:Fi,

11:34 but they extended that

11:34 to campus and branch networking for wired LAN.

11:37 We extended that to SD-WAN,

11:40 and we also are now extended that

11:41 to a security access points such as SRX, VSRX and so on.

11:45 So together you get a cloud managed,

11:49 AI-driven, automation based operations,

11:53 which are allowing you to develop

11:55 and deploy self-healing networks.

11:57 So if you're a service provider, you take all that bundle.

12:00 You can start offering that to your enterprise customers,

12:03 to accelerate the digital transformation,

12:06 and still be secure,

12:08 and be very agile with respect to and efficient

12:12 with respect to how you operate the network.

12:14 - Raj, of course, we can't ignore

12:17 the two experience enablers that you mentioned,

12:20 that cross all of the three solution pillars.

12:22 And that's connected security and intelligent automation.

12:26 Maybe if we start with connected security,

12:29 give us your perspective in terms

12:30 of our approach to security.

12:34 - Sally, as you said,

12:35 security is very important to our customer.

12:37 Service providers want to make sure that infrastructure,

12:40 as well as the managed services

12:41 they offer to the enterprises is secure.

12:44 But many times they are worried,

12:46 that introducing the security

12:48 such as advanced threat protection, malware detection,

12:51 will slow down traffic,

12:53 and that will impact the performance SLAs they offer.

12:56 So one of the things we invested is

12:58 in insecurity should not be an afterthought.

13:01 It should be integrated into a product,

13:03 but more importantly,

13:05 security should continue to operate at line rate.

13:08 So we have invested in our custom chips,

13:11 as well as software to make sure

13:13 that any security mechanisms we introduce

13:16 for advanced threat protection,

13:18 encryption and so on, simply works at line rate.

13:22 Our customers service providers get the best possibility

13:26 with respect to being able to offer secure networking

13:29 to their customers.

13:31 - That's great, thank you, Raj.

13:33 And then of course,

13:35 the second experience enabler also important.

13:38 That's intelligent automation.

13:40 If you could maybe expand on our approach there as well,

13:43 that would be great.

13:44 - So I think Sally, as we discussed earlier.

13:47 One of the two experience enablers

13:50 for our three pillars we mentioned,

13:52 is intelligent automation.

13:54 And we have been investing heavily

13:55 in many aspects of intelligent automation.

13:59 Let me start with one example at a time and expand on it.

14:02 So first of all, our Paragon automation suite,

14:05 where we have invested heavily to automate

14:08 all the aspects of network design, capacity planning,

14:12 day zero installation, day one operations,

14:14 day two upgrades and so on.

14:16 And while doing that,

14:17 we're applying artificial intelligence

14:19 and machine learning techniques,

14:21 so that we can do AI-driven predictability,

14:23 AI-driven insights inferences for closed loop automation,

14:27 and self-remediation kind of experiences.

14:31 In addition,

14:32 we have moved our overlay networking solution

14:34 to cloud native,

14:35 where again, we provide automation

14:38 in terms of both a control plane,

14:41 as well as life cycle management.

14:43 Apstra, as I mentioned earlier

14:45 is our fabric management solution,

14:46 which is a multi-vendor,

14:48 but more importantly, it's intent-based.

14:50 So we specify the intent

14:51 in terms of how the network should look,

14:53 how it should operate,

14:54 and Apstra makes sure that by continuous monitoring

14:57 it make sure that the network does not deviate

15:00 from the intent, if it deviates it corrects itself.

15:04 Our Mist AI, and Marvis engine is another great example.

15:07 Using that we are providing automation across

15:10 all aspects of cloud managed solutions,

15:13 for Wi-Fi, wired networks, SD-WAN, security access.

15:19 And finally, last but not the list,

15:21 is our investment in RAN Intelligent Controller

15:25 for 5G ORAN Networks.

15:27 RAN Intelligent Controller is the new operating system,

15:30 or the brain of the Radio Intelligence,

15:32 Radio Access Networks.

15:33 Using the RIC and AI-ML based xApps,

15:36 they are trying to provide automation with respect

15:38 to how you optimize the spectrum used by the RAN,

15:42 how you optimize services are provided by RAN and so on.

15:46 So the combination of whole variety of things

15:48 we're doing at Juniper are all contributing to this stuff

15:51 you call intelligent automation.

15:54 - So let's sort of change gears a bit

15:55 and talk about openness, Raj.

15:58 So, our industry is pivoting pretty hard

16:02 to multi-vendor collaboration,

16:04 dis-aggregation, open ecosystems, et cetera.

16:08 And to be honest with you, that's been pretty much

16:10 at the center of Juniper's history as well.

16:14 And we've led all of the above.

16:16 What are we doing these days to continue that leadership

16:20 to continue to remain relevant.

16:22 And maybe talk a little bit about

16:24 some of the standards bodies

16:26 that we've been particularly engaged with

16:28 to move this along.

16:31 - Yeah, so I just said like Juniper has been a challenger

16:34 in the internet industry, right from the start.

16:37 And if you look at IETF,

16:39 which probably sets the internet standards the most,

16:42 Juniper has been at the footprint of

16:45 the open standards developed by IETF,

16:47 starting with MPLS, RSVP,

16:50 then recently Segment Routing,

16:53 all of the network management standards,

16:54 let's say it's OpenConfig, and NETCONF and so on.

16:58 So I think it's in our DNA

17:01 to encourage open ecosystem

17:05 because not every player or each single player

17:08 cannot provide all the parts,

17:09 and having such an open ecosystem allows us

17:12 to fairly compete and still continue to innovate.

17:16 We have done similar things in other areas.

17:17 For example, the Linux Network Foundation,

17:20 we have been one of the leaders

17:21 of software-defined networking standards.

17:23 We have been recently been contributing to their blueprints

17:27 for edge cloud networking, right?

17:29 We are doing more and more in the 5G ORAN,

17:31 which is the alliance formed by Telcos to create

17:34 very open ecosystem for Radio Access Network.

17:38 We are leading working groups there in the area

17:40 of transport slicing,

17:41 in the area for many interfaces for RIC,

17:45 as well as as xApps and so on.

17:47 So I think we'll continue to,

17:49 and I think another example I want to give

17:51 the open ecosystem

17:52 is that our Apstra based fabric automation.

17:55 It's the first ever open fabric management solution,

18:00 that spans multiple vendors.

18:02 It's completely vendor-agnostic.

18:04 It allows you to build a closed loop automation

18:06 for a three stage, five stage growth fabrics,

18:09 independent of each vendor is involved.

18:12 Why do we support that?

18:13 Because we really believe that being open,

18:15 we make the customers have a very sort of easy access

18:20 to any vendor while taking advantage

18:22 of Juniper's software innovation.

18:25 So I think we'll continue to do that,

18:28 and we do believe that encouraging

18:30 the open ecosystem system is good for us

18:32 as well as everyone else.

18:33 - So Raj, really great in terms of the details around

18:38 the three solution pillars,

18:39 and our experienced enablers with security

18:42 and intelligent automation.

18:45 Maybe if you could talk a little bit about

18:47 how these solutions truly deliver outcomes

18:50 that matter to us service provider customers,

18:53 and their customers, based on some of the market dynamics

18:57 we discussed earlier in the conversation.

18:59 - So as I said, Apstra based fabric automation

19:02 that we are delivering today.

19:04 It provides you intent based framework,

19:06 by which you can specify

19:08 what your network fabric should look like,

19:10 and how it should operate.

19:11 And now using closed loop automation,

19:14 like continuously monitoring the fabric,

19:17 Apstra to make sure that the designed intent

19:20 is actually followed, and any deviation

19:22 from the intent is automatically correct.

19:24 So that's one, similarly in the area

19:26 of secure and assured experiences I talked about,

19:29 we have no actual data points.

19:31 For example,

19:32 the Paragon automation suite I talked about

19:34 has one past called Active Assurance.

19:37 That is now able to reduce

19:39 the number of false service deployment rates by almost 13%.

19:45 In the area of security, for an example,

19:47 there's a independent lab called ICSA Lab,

19:50 which has done a study of our malware detection,

19:53 and found that we are able to detect malware in traffic,

19:57 100% of the time without any false positives.

20:00 That's an incredible result you would not expect.

20:03 So I think those are the two areas

20:04 where we can say we have active outcome based results

20:08 for secure and assured experiences.

20:11 Now, if you talk about efficiency,

20:12 we talked about in the context

20:13 of cloud-based infrastructure,

20:15 people are expecting efficiency, agility, resiliency.

20:18 We invest heavily in custom chips,

20:21 so that we can deliver much better performance,

20:24 much better security and so on.

20:26 So you're getting the six-fold increase in power, energy,

20:30 power performance, and efficiency,

20:33 which is important for the environment and being responsible

20:36 with respect to green computing.

20:38 Finally what we have done is that

20:40 we have built-in security because of our custom chips.

20:43 So for example,

20:43 we support we support MaxSec and IPSec.

20:46 So MaxSec allows you to do Layer 2 end-to-end security,

20:49 including encryption.

20:50 So together, I think this is an example

20:52 of how we can also not only delivering high-performance,

20:56 assurance, security,

20:58 but also on demand efficiency with respect

21:01 to being responsible for their environment.

21:03 - Fantastic Raj,

21:04 again, thank you for your time.

21:06 It's always a pleasure to chat with you,

21:09 and for everyone listening in,

21:10 listen, if you wanna know more

21:12 about Juniper service provider strategy,

21:15 I invite you to go to juniper.net\serviceprovider,

21:19 and you'll find a little bit more information

21:21 about what we've spoken about today.

21:23 And thank you, thanks everyone for listening in,

21:25 and Raj, thanks always for your time.

21:28 - Thank you, thank you, Sally.

21:29 (cool music)