Expanding API Developer Portfolio – Reza Shafii, Kong

This is Textron TV. I'll have the great pleasure being joined by Reza Shafii Reza is VPR product with Kong. Welcome Reza.

Thank you. Great to be here Mitch good to be talking with you when we jump into some of the great news that from announcements you've had recently before we do that tells about yourself. You have an impressive background great background until course.

What Cong does? Yeah, for sure. So resident shopping VP of product at Kong is she said and I'm being here for three years my general background salmon.

I call Api and middleware geek was at ba systems back in the days and then Oracle and then meal soft core OS red hat and now here at Kong. So what caused us is next Generation API management, he can't imagine us, you know has been around for a bit. Now.

We have a cloud native full life cycle API management stack that we're very proud of and that we announced a lot of capabilities around last week at our Summit, which is our customer. Fantastic. Well, let's get into that.

So one of the big guy announcements you just mentioned was around the API Gateway, which I think is in GA now right publicly available. Tell us a little bit about that. I mean this has been in a works for a while.

You've had it in Tech preview and beta I imagine and now getting it in you know for sale customers hands tell us what's happening with the product maybe with some of the unique things are about it. 0 is our first major release. We'll follow us semantic versioning models.

It's the first major release in 18 months. So it's a pretty big deal for us. And the reason for it is that not only it provides major new capabilities, which I'll talk about in a second.

0 as well in terms of the functionality, you know that I divided up in my mind around the open source version and the Enterprise version one open source version. We have introduced two things that I would say are really not worthy one is you know, I Gateway enables you to effectively create routing rules and take those rotting rules and map them to back and upstreams and in the middle inject some logic that I need to protect the traffic or redirect to traffic or detect, you know intrusions and so on. Our routing rules have significantly expanded now in terms of being able to have a whole new domain specific language for specifying those routing rules.

It's not just that regex anymore. It has and an ore and combination of those things which makes it a lot more flexible for creating routing also really excited about that. Also, we've introduced an OpenTelemetry plugin which enables us to work with any OpenTelemetry compliance monitoring solution literally click up a butt.

Right like you put the plug-in say, this is dated dog. This is up Dynamics. When you OpenTelemetry compliance vendor or tool and now you get the tracing capabilities around your and you know, you can apply that plug-in at any route at any Upstream globally, you know, the oliflexibility of the Cong Gateway.

So those are the two really exciting things at the open source. Level there's a whole bunch of stuff on the Enterprise. I shall get into but any thoughts Mitch on the on the open source stuff.

Yeah, I'm really intrigued. Well great to hear. I mean, I love regex super powerful but not accessible to everyone, you know, he could spend a career learning regist and tricks and tips.

I'm really interested in your your plug-in architecture because that's sometimes is the part. That's a little tricky wonky to create and do it. Right and it takes it.

Sometimes it takes multiple tries to get it right that sounds like with your your OpenTelemetry and some of the capabilities you have and plugins. You have the ability to connect with the course of the third products, but I imagine you're also talking about kubernetes as an environment that you're operating with and within and interconnecting with that. Is that a fair guess on my part?

Absolutely. So you touched on a couple of points. I think that are really interesting one is You know one of Kong's secret sauces is plugging compose, which is dynamically you can inject this pieces of logic within different parts of the plug-in lifetime of the request and response lives.

Mmm. And that's interface that players is open. So there's thousands of plugins out there that have been developed.

Usually when users or customers ask me. How can I do X. Is there a plugin for that?

I just do a Google Kong, you know, whatever and then you see someone has built the plugin for it and we have over a hundred that we provide audit a box. The flexibility of those plugins now is is you know respected routing rules is it's not changed that you can you still apply the plugins at the same level but the routing rules decide how the actual requests are mapped to the routes that you've created but you bring up an interesting point which is in the Enterprise version with introduce something called plugin ordering one of the top requests from our customers was I applied this plugins, but can I decide plugin a runs first, you know, maybe always do talk and validation with my customer, you know server and then after that Performed for the following request transformation right before you couldn't have that level of flexibility on how their plugins are ordered now it is there, you know Enterprise. And then you touched on kubernetes.

com Gate 1 kubernetes uses admin interface, which is the restful API and you use I told we got called deck for for devops, right which basically does declarative config with Gateway or you could run it as an Ingress controller. 6 which supports the kubernetes Gateway API in data, which very excited about and we let the charge in creating them. So until you could apply crds to apply things like HTTP route and so on but now you apply those standard kubernetes.

So we're standardizing that layer on kubernetes, which is very exciting. I'm interested curious about that. Is there certain scenarios where the Ingress approaches better or more capable than the other approach and tell us a little bit about why both and why that's significant.

Yeah, you know, it's interesting because this is a question that comes up a lot from our customers as well. I think really it comes down to your operational mode of preference right? There are many customers who increasingly like to operate their kubernetes clusters using crd's alone, right and map that to things like get off schools right that that work on this crd level and in those models the Ingress way of doing it is the right way of doing it.

On the other hand, there are customers. They go kubernetes is just another deployment Target and I actually want to have another layer on top of kubernetes that standardizes the way I do config management deployment and for those customers using our standard interfaces better because they have non kubernetes and kubernetes ways and they just go it's all it's all good homecoming it's not kubernetes. I don't take a side on this tree.

The customer is always right we can do both I can see distinct advantages like do I want to venture into kubernetes and manage it through you know that path or I'm comfortable in Kong. Let me do it this way and do all my things here. It makes sense.

Good. Well talk some more about some other announcements from last week the approximate about when when not sure when this will air but I mean, we're good to away from your from your summit. Yeah, so we also announced the slew of capabilities around Kong connect can't connect is our Cloud solution added.

com and literally within a click a few clicks of the button. It enables you to run a common Gateway, right? Then we got a Plus Credit Card based here as well as a free tier as well as Enterprise here.

So come connect has been really taking off for both sort of single Gateway users because it takes away the operational complexity of running the control plane. So just like the easiest way to run con Gateway, but also more and more and we had a lot of great customers come in and talk about this there is this platform Builder pattern in organization. That's I'm sure you were The people who build a platforms that make it available to all the app Builders and they have more and more for a need for something.

I call connectivity as a service. They're basically building a connectivity platform and connect provides a one button connectivity platform that then their team can self serve on can come in and provision their gateways within clear swim lanes that have proper R back and then have a single catalog of all those services and on top of that then they can product us those API search them how full Analytics Connect us got a number of new capabilities. We announced around multi-geo support around integration with AWS Marketplace access to connects API Straight Talk and based authentication model this concept of runtime groups.

And it's just the list is too long, but we introduce a whole bunch of stuff on connect as well. Fantastic, you know one question I've had about your product configuration. How does mesh fit into this is API Gateway in that capability kind of foundational to the mesh or as mesh another product on top of this.

How does that fit together? Yeah, it's a great question. So.

Philosophically on our product line we've got we've got four products and connect which brings them together. We got the con Gateway County Ingress controller, which we talked about Kong mesh, which you're bringing up calling insomnia, which is our death tool and connect which brings them all together. and our philosophy around this stack is that it's an incremental buying model.

We do not want to push the whole platform down. Folks just throw outside. We do just don't want to say you need to do the whole thing all at once or it's not all or nothing right?

You can start with any and then progress together example. He can start with the mesh for services connectivity management. And our mesh is really designed to make it easy to use and work with both kubernetes and nonkipedia.

That's really why we created our measure. But then no mesh is an island, right? It needs to have an Ingress.

And guess what the con Gateway Ingress works really well with the common mesh Solutions and just plug and play that and take advantage of the concave way. Now if you suddenly have multiple teams doing that and you want to do that platform building stuff, you can tell all your teams connect those gateways to connect and then now you have single pane of glass visibility across and try to put our back. Same thing goes with our insomnia, right?

You can start with insomnia to build apis consume apis, but now you want to do get AppSec for API specification, which is something we call Api Ops. Well, you can generate from insomnia you saw a plugin and generate declarative config for our Ingress controller or for our Gateway. So it's just all meant to work together as pieces of the puzzle.

And you start wherever you want make sense. I can see how it fits together to what is the most common path. Where do you find people that are You know in the process adopting Cloud native and microservices and maybe they're maybe they've gotten their heads around service mesh yet.

Usually where do they start when they start working with Kong? Yeah, so what we see is. Two very sort of clear bottom-up motions.

I would say is and users typically will need a Gateway, right? That's that's a very clear one. Like I've built an app or I've got a set of AppSec that security is missing.

The right thing is missing. I answered some custom Logic on top of this to integrate my IDP whatnot. So they start with the gate that's what the other is developers need to just call some apis.

Right? And this is this is the an area that is fascinating to me because Carl just not enough they need to do collection of calls build tests. That's another story.

That's insomnia got the Gateway has to start More and more we see mesh Builders like developers really administrators who was like, I need zero trust security. So I need a mesh that's also one but but more often than not mesh comes in as a Evolution after you've had Gateway installed and you figured the Gateway side of the house now, so it's more of an evolution like and then connect is really says how to run the whole thing much faster. Okay?

Fantastic. That's super helpful. Well tell us a little bit.

I know you have sort of an incubation process that you bring Technologies through before they come out to General availability. What are some of the kinds of things that you have in the works? com.

And you will see a number of projects that have been launched there. One of them is called Vasan X which enables you to program anything in the lassum supported language like JavaScript golang's rust c++, and run Ms. Complex effectively, right?

So that's that's really cool. It runs at Native in process speed. We introduce in you operator, which does you know, D2 Automation and management and and you Kong Gateway control plane that's entirely in goaling and that seems like managing multiple versions of Cong Gateway at once or integrating very easily with different databases.

So that was called Coco everything at Kong has to start with k, of course, so lots of interesting stuff on the incubator side. Yes. Fantastic, it's great to get exposure to things that are in the works and get an eye towards where it's heading.

So do you do your summit once a year or is this something you do more frequent than that? Once every year that's that sort of events in San Francisco. Yeah, and it's it was just great out.

The energy is still there. Oh, I feel we had enormous attendance and lots of lots of good customer stories as well. Very good.

I imagine you're gonna be a coupon. We'll see you there. I'll be there of course.

Yeah, we'll have a big present stand also a wrestling event. Those are the events reinventually great. Yeah, absolutely two great conferences coming up just around the corner here.

Well, it's been fantastic talk congrats on the on the announcements. Especially you know that big of a release, you know, something been working on that long. It's nice to get that sort of out the door and in customer's hands and know that it's going to be going into production environments with customers.

So great talking with you. And I hope you come back and tell some more about Kong and new things that are happening and and special sharing. What customers are experiencing what some of the challenges that they're seeing is their adopting Cloud native.

We're all looking to learn from each other, and that's a great thing about both the vendor Community Technology Community as well as end users practitioners. So so folks can learn more about Kong and you're offerings. com.

Our website has everything you need and yes Mitch. Thank you very much for having me and it's been a pleasure. Absolutely.

We'll see you again soon. Thank you.