Kong Kong Summit 2020, 5 Nov 2020

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Customer Stories, Insomnia 2020.5, Service Hub in Konnect, Kong Mesh 1.1 | Kong Summit 2020 Keynote

Description

Kong leaders, along with industry special guests, will continue to explore connectivity in this new digital reality and unveil more exciting, new products that will redefine what connectivity means today — and beyond.

Learn more about Kong: https://bit.ly/2I2DypS

A

Thank you to all of our khan summit, 2020 partners and sponsors. Let's welcome our vp of marketing sandeep singholi.

B

Welcome to day two of the kong summit as a marketing leader. What excites me most is the positive impact kong is having on its customers throughout the conference you heard from several of our customers on how their relationship with kong brought out the best in them today and right now. I want you to imagine how kong can bring out the best in you and your company.

B

Let's look at three stories about three strong companies made stronger through their relationship with kong.

B

The first one is verifone. If you have a credit card or a debit card, you certainly use one of their more than 35 million payment terminals around the world.

B

While verifone is the largest payment solutions, provider in most regions, and they saw a need for them to develop solutions for the e-commerce segment, there were also new entrants in both e-commerce and point-of-sale landscape. Competition was increasing.

B

Verifone's unique value proposition was the relationship that they already had and a global payment system that they had built, but this global platform was in fact 20 disparate regional systems.

B

What verifone needed to do was build a consistent global platform and also build a only channel solution that catered both to the point of sale, use cases and the e-commerce use cases.

B

This is where com helped them verifone leveraged kong enterprise as a service connectivity platform to centrally manage and secure the entire globe's traffic.

B

They moved their monolith, applications to a microservices architecture, with the flexibility to manage their apis on premises in private cloud, in public cloud and in hybrid deployments they also abstracted their apis for the end user and because of kong, they were able to provide standardized api documentation using open api 3.0.

B

The impact was huge. They were able to reduce the total cost of ownership by 75 percent and decrease time to market by 20 times we helped verifone with their digital omni channel strategy. Imagine how we can help you now. Let's talk about another company that we are proud to be working with cargill, the agro business giant they operate in 125 countries and handle the logistics for over 33 percent of the food world's food supply.

B

It's a huge responsibility when you are responsible for the nourishment of one third of the world's population and cargill always realized that they need to be constantly innovating.

B

However, they saw their I.t systems as a bottom leg for them to roll out these digital services rapidly. This was because these systems were siloed, which made it very difficult for them to maximize resource efficiency.

B

Kong worked with the amazing platform team at cargill to help them build a unified developer experience across legacy and cloud native systems.

B

They built a platform exposing apis for both the internal and external developers, with kong's flexible plug-in architecture. They were able to support the varied use cases.

B

The this enabled the rest of the teams to build services that the entire company could reuse.

B

Furthermore, with kong, they were able to develop this platform in a box that was compliant and secure and give it to these teams for them to develop their applications and services rapidly and independently.

B

Because of that, the development teams were able to focus on creating services that had a business value rather than wasting time in compliance and pipeline and other policy related issues. The impact was huge. They were able to achieve 65 times faster deployment and create more than 450 services. In just six months, we helped cargill create a digital platform.

B

Let us help build yours.

B

Verifone invited us to help them with their omni channel strategy. Cargill gave us the opportunity to create help create a digital platform. Now, let's talk about vestas westers is the largest manufacturer of wind turbines in the world.

B

They have installed wind turbines in 81 countries across six continents. They are huge, but the challenge was: how do they innovate beyond what they already had? They saw an opportunity to leverage data for that next level of transformation.

B

This is where they are leveraging kong. There are two teams that are working with us from vestas. The first team is the power solutions group that is responsible for the sale, service and maintenance of the wind turbines.

B

What they're doing is very interesting. They are abstracting the data which was previously just used internally and providing api interfaces to that data for the other teams to build value added applications on top of them.

B

One example is the what what they're doing for the customers uh to decide where to install new wind turbines.

B

You see these wind turbines run into millions of dollars, and what vistas is doing is they are triangulating data from their own data sets with whether agency data and the data collected from the installation of the customer wind turbines and helping figure out for the customer where they should be installing this new wind turbine for maximum power production.

B

The second team that kong is helping here is the global I.t team. They use kong for their internal platform that provide apis and services to both legacy teams and those working on bleeding-edge kubernetes environments.

B

They run kong on-prem in cloud and also in kubernetes environments, because of this great work, vestas is able to build towards a future of providing this valuable data externally using apis, which they had previously just used internally.

B

This is helping them to provide more value to their customers and potentially new sources of revenue for the company. The company is also changing a culture from a traditional central I.t to a data centric modern software team, where apis are used by the different departments to talk to each other.

B

The company is moving from being a turbine company to being an iot data company. We are helping vestas towards a digital data future. Let us help you build towards yours.

B

Kong is here to help you be the best you're, all strong companies, partners. Friends, let us help you shine. These were just three stories. What I want is you to be here next year. Talking about your story on how kong made you successful, download our case studies, take it to your leaders and tell them that you want to build this with kong as a partner.

B

Thank you. Let me now invite our vp of products reza shuffy.

C

Thank you sandeep and hello. Everyone, as you may remember, from yesterday's discussion. One of the central end users of connect are developers.

C

And kong's answer for accelerating developer lifecycle is insomnia.

C

Now insomnia is a tool and at the end, what really matters is the workflow that the developers use to accelerate their end-to-end life cycle for service management, and here at cog, the end-to-end flow. One of the main end-to-end flows that we focus on is the following.

C

We believe very much that the spec needs to be the source of truth for the api's signature, so spec based development is really critical to us and when I say spec based, I'm being very deliberate there. Sometimes the space the spec comes to being after the quote, for the service has been written. Sometimes it comes to being before it doesn't have to come first, but what's important is that the spec be declarative, config, driven an open api spec for restful. Apis, of course, is the primary standard.

C

That's been gaining traction and used primarily out there now. The flow that we we see often our customers use, is that they use the spec, and then they make sure that that is interjected with policies around security and reliability. That then need to take effect appropriately on the right target environment.

C

Now these policies themselves need to be stored in a declarative, config model so that their change can be managed as a flat file, like any other change through the good old pr model and reviewed in that way, and then applied verified and applied through a cicd automation pipeline.

C

And for that to happen, not only your run times need to be declarative. Config enabled you need the right tooling, at the development lifecycle phase, to make sure that you can deal with multiple protocols to make sure you can inject the policies into the right places within your spec and generally collaborate with others.

C

Now, as I said, insomnia is the tool that we have been investing in quite a bit to enable this type of workflow.

C

Now without further ado, let me introduce paolo morello who's, been introducing insomnia in his company to enable developers to achieve higher productivity.

D

Hi reza and thanks for inviting us yes, abhijob, is using insomnia since the beginning, and it has a game changer uh for our productivity uh insomnia. It's so easy to use that even our non-technical team is using for daily operation. uh They use it for a cool internal api, and this was great because uh our our coders uh didn't have to create uh and a web panel for that. uh So we saved a lot of time. uh Talking about uh how easy is is, is working on uh on within some app.

D

I can't say that our support team uh go crazy for the c url um copy and pasting a feature that permits them to recreate very complex http called uh inside insomnia, and this is saving us a lot of time, because http calls are very complex and insomnia is recreating them uh very, very fast and it and it's it's very easy.

D

As I said, both our technical and support team uh use, insomnia with for their daily tasks and one of the most appreciated function is the mood environment.

D

This permits us to share the environment and shed switch between our beta and our production environment uh with just a click.

D

As I said, our our team is distributed uh between rome, london and and boston and the so the work, the shadow workspaces and the sync function, uh it's very precious for us, because um these features permit us to stay in seeing and collaborate uh with success um yeah this. These are the great features that we are using here in in epidrop and that's why we love insomniac so back to you reza.

C

Thank you, paulo, that's great stuff.

C

Now it's been a year since we announced the acquisition of insomnia, and quite a bit has happened since then.

C

Not only has the product really been popular with the community in terms of usage and in terms of contribution to the project itself, we've also been introducing quite a bit of capabilities over this timeline, starting with the introduction of design capabilities, a plug-in hub that allows accessibility of insomnia with various plug-in again is contributed to by the community quite a bit in so the cli that enables further automation, around insomnia, based service lifecycle, automation and, just recently, the capabilities to build unit tests for apis in insomnia itself.

C

Now let me introduce the jiko and melissa.

E

Thanks reza and what a great story from paulo huh as well as api drop, we work with companies of all shapes and sizes, but something they have in common is the need to support loads of different types of messages now not just restful apis, but with this comes the inevitable challenge for developers of having to manage and switch between more and more tools to make this happen.

E

This isn't really much fun. To be honest and, frankly, we want to spend our time on the fun part, building the services rather than getting bogged down by all the different tools we have to use, and this is why we've got something really exciting to announce.

F

Introducing insomnia2020.5 with grpc.

E

Awesome so no more switching between api clients anymore. Can we see it.

F

Yeah, let's show everyone what we've been working on on the insomnia team. Let's start with the basic grpc request, like a unary call where you send a single request and you get back a single response if you're familiar with rest, this will look very familiar to you.

E

But, unlike rest, grpc cuts down the message size right, so it's really popular for microservices.

F

Right, we only send the essential data which helps with scaling. So let's get started, let's go ahead and create our request.

F

I'm going to create a request, and I'm going to name it say hello and I'm going to choose grpc from the drop down.

F

Insomnia is then going to prompt me to select a proto file, and this file tells insomnia what services and methods are available. I've already preloaded one into insomnia, called hello service, so I'm going to go ahead and select that and continue, and now I'm going to go ahead and select the service and method here from this drop down. I'm going to choose, say hello and insomnia takes that profile builds the body for me and tells me also that it's going to make a unary call.

F

So let's go ahead and start this request before we can do that. I need to pre-fill the url and I'm going to be using grpc bin, which is something similar to mockbin all right. Now, I'm ready to send this request. Let's go ahead and try it out and there we have it. Our first grpc request in insomnia.

E

Hello and if you wanted to change the request, so you can test different responses, it's a really easy change in the ui there right.

F

Yes, exactly it's great for rapid iteration now, admittedly, this isn't the most exciting request that we could make. It's just like a rest request where we're sending a single request and getting back a single response. Let's try something different something everyone might not be familiar with something unique to grpc like streaming. Maybe a client streaming requests. So let's go ahead and try one of those I'm going to go ahead and create a new request.

F

I'm going to call this lots of greetings. Grpc is pre-filled for me, so that's great! I'm going to go ahead and create that, and now it's asking me for the same protofile I'm going to choose the same one as before hello service, I'm going to continue and this time I'm going to choose lots of greetings from the drop down.

F

Now there are some changes here. The body is still pre-filled for me, but it says client streaming and it also has changed the send button to start streaming so before I can start that stream. Let's go ahead and enter the url, we're going to use the same urls before grpc bin, and now I can actually start the stream. So let's try it out.

F

Perfect insomnia has actually created my first stream for me, which is like a queue of requests that it'll send over to the server and get back a response, so I can actually add more to this stream. So let's go ahead and do that it's added one more next to the first one, I'm going to add another one. Just to you know, try it out fully so perfect. I think three is good enough. uh Let's try this out. Let's try! You know sending this over.

F

So I'm gonna commit this to the server all three requests here and perfect. There we go. We actually got back our three greetings here from the server uh in a single response, so this is client streaming uh in grpc.

E

So many hellos- what's also cool to point out here, is how your requests are being saved down the left-hand side there. So you can actually store this as a collection like hello, grpc and then go back to it anytime. You need.

F

Yes, exactly it's, we actually use this inside the insomnia team for administration apis which are made in graphql, but now we can do this for all of our grpc apis that we build out as well.

E

So because you're sharing these collections across your team, it means you can get things done more quickly right, because everyone in the team can just get stuck in with the collection and use it to start making requests immediately.

F

That's why insomnia2020.5 is such a big deal for users?

F

It's the only tool that supports rest, graphql and now grpc, opening the doors to supporting even more uh protocols like websockets, for example,.

E

Which, obviously, you can manage in the con gateway too, but it's worth noting that, like with all your services building and managing them, is only half the story. I mean they're only valuable if they're actually used.

F

Back to you reza.

C

Thank you, nietzsche and melissa. Let's switch gear a little bit and talk about the enterprise, architects and service hub as well as portal.

C

Service hub and portal are really two sides of the same token. Service hub enables a system of record for all of the enterprises, services or all of an organization services. It is the single place where services can be categorized easily searched for socialized within the enterprise.

C

Dev portal is the other side of the token for services that require it.

C

They can be published to a dev portal where they have a formal page for discovery with their documentation and a formal application registration process by developers that want to consume the service can be used.

C

So in some ways, if you think about it in terms of a store analogy, a service hub is like the inventory, whereas the dev portal is the front end of the store, but both of them are critical.

C

Now let me introduce stephanie and jordan to show us the service hub and portal in connect and some of the end-to-end workflows that one can.

G

Achieve great, thank you reza today, we'll be looking into how the enterprise architect can use connect service hub to manage and publish documentation to the dev portal, as well as how developers can leverage application registration to request access to those services.

G

We'll start off here on the service hub as reza showed us. The service hub is a system of record for all services in the enterprise.

G

You can see how some services have been published to portals the portal and some have not we'll go ahead and click into the account and create service.

G

So this is the service overview and you can see that there is a single version before we look into that we'll go into this new documentation, tab in connect. We allow you to document each version via oas specifications and now we're introducing support for markdown on the service level to provide more flexible documentation options across all versions over time, we'll be adding first class support for other specifications such as async api.

G

Now go over to the versions tab. Here you can see the overview for the version we've added in a new spec tab once you've uploaded your spec, you can view it for reference and it's a preview for how the developer would see it in the portal.

G

Now we'll go and publish this service to the portal, so using this actions menu up at the top right here, you can simply publish it to a portal. This will publish all versions to the portal for application registration.

G

Any versions that have been deprecated will not be available for app registration. Now I'll pass it over to you, jordan, to view it what it looks like in the eyes of the development.

H

Thanks steph, we are currently viewing the service view page in the developer portal. You can see both the service-wide markdown documentation, as well as oas specifications across versions. From this view, currently you are seeing our dark theme preset. The portal not only provides easy theming options, but full html, css and javascript customization.

H

If we click click to the portal catalog page, we can see a table of all services published to our portal. Here you can search across the services and their versions as well.

H

Let's visit the my apps tab to look at app registration and connect.

H

Here's a list of applications that I have created as a developer applications allow developers to access protected services like the accounts inquiry service we published earlier.

H

Let's go back to the accounts, inquiry service and register one of our applications against it in the upper right hand, corner of the service view page, there's a register button.

H

When clicked, we can select from a list of our pre-existing applications to register with this service because it is protected. This developer will need to wait for administrative approval to access this service.

H

If we return to our my apps page and click on the my accounts, mobile app, you can see that I have a now have a pending registration request against the accounts inquiry service.

H

I will receive an email once approved after approval. I can use the authentication credentials above in this case oauth2, to authenticate against the proxy that wraps up our preview of how the service hub integrates with connect portal.

H

We saw how enterprise architects can expose different types of documentation for published services to developers via the portal and how application registration allows developers to request access to those services with that I'll pass it back to reza.

C

Thank you stephanie and jordan. This was a great demo now today, I'm very excited to announce the availability of kong. Mesh 1.1 beta kong mesh, as you know, is the product based on the kuma project upstream, the universal service mesh. Let me tell you a bit more about the capabilities of kong mesh.

C

We start with multi-mesh support and this is critical, because what it enables you to do is create separate scopes of control for different sets of applications say you have a set of applications that could be running on their own data center on their own cloud on their own kubernetes cluster, that you can designate as one mesh and another set running on their own context, and then you can design, define different policies and target them to these meshes separately.

C

But yet at the top level, congmesh has a single control plane that enables you to achieve full single pane of class control over all of these different meshes, and this enables enterprise architects to enable different teams to have the flexibility to have their own infrastructure choices, yet be able to have a single place of control over all of these different meshes.

C

Another really exciting capability that kong mesh introduces is vault support. Hashtag vault is a very popular solution that we see out there for managing, not just certificates but secrets in general as well, and kong. Mesh comes with out-of-the-box integration with vault so that it can leverage the certificates from a vault installation in order to use for its policies and then last but not least, load balancers are obviously critical to the interworking of service to service communication.

C

But with the measure in place, we are seeing more and more the ability to actually load balance between services purely using the mesh and kong mesh comes with native capabilities that make this type of routing and load balancing much easier.

C

So this not only achieves higher efficiency, but it also achieves a degree of effectiveness because it lowers costs by removing the load balancer and also removes the load balancer as a single point of failure and a single point of traffic.

C

So this is a really exciting use case that we're seeing and which we're going to be investing in quite a bit. More all right, well, certainly has been an exciting two days. Let's take a look at some of the things we've covered and what it all means we began by talking about the big changes that are happening in the technology landscape, but also in the way that users consume connectivity software and we talked quite a bit about connect, kong's full stack connectivity platform. That's designed from the ground up for the cloud native era.

C

We introduced all of the connect components, the run times and the functionality on top of it, that's delivered as a service, and we talked about you, the end user, and how connect is the platform designed with the end user in mind and for the end user era and importantly, in the last two days we talked about how here at kong, we believe in all kinds of connections, reliable anytime for everyone hanoho. You can count on us to continue to innovate in the connectivity space to make this vision happen.

C

Thank you for spending the last two days with us. I would also like to thank our partners and sponsors. Thank you for your sponsorship of kong.

C