EdgeX Foundry’s v1.3 (Hanoi) is the seventh consecutive semi-annual release, which features simplified deployment, improved performance and scalability testing, and an updated command-line interface (CLI). Hanoi also incorporates the first collection of new, platform-wide micro service APIs that allows adopters to get a feel for what’s coming with EdgeX 2.0 in the spring. To dive deeper into the release and other aspects of the project, we spoke with Jim White, Chair of the EdgeX Foundry Technical Steering Committee.
Here is the lightly edited transcript of the interview:
Swapnil Bhartiya: Hi, this is your host Swapnil Bhartiya and welcome to TFiR Newsroom. EdgeX Foundry recently released Hanoi, which is the seventh consecutive semiannual release. And it has a number of features including simplifying deployment, scalability testing and launch of command line interface. It also incorporates the first collection of new platform-wide microservices API, that allows the users to get a feel of what’s coming with Edgex 2.0. That’s the most I know about this list. To dive deeper into the release, we have with us, Jim White Chair of EdgeX Foundry TSC. Jim, first of all, it’s great to see you or hear you again after a long time. I think last we talked was in September during the hackathon.
Jim White: Swapnil. It’s great to be with you again and with your listeners. And Yep, unfortunately, 2020 I think is a year of lost chances to chat with people. So good to be with you again.
Swapnil Bhartiya: The new release is here so tell me a bit about the release and the highlights.
Jim White: Sure. This is, as you mentioned, our seventh release. So we’re pretty proud of that, in that for seven consecutive cycles. Now we’ve been able to release pretty much on time, this is a dot release. What that means is that there are a significant number of features, but it should be backward compatible with everything up to what we call our Edinburgh release, which was our one Dotto release. This will be officially one dot three things that are coming out with this release including the beginnings of a new API set. So for those who are watching Ajax, we’ve had the same API for three plus years. But there’s a new purse, new API set on the horizon. And this particular new release has a subset of that getting ready for a tool release in the spring. Additionally, we’re providing the ability for device services to live on a different host and more easily connect to the rest of regex. We’ve added some new features in terms of support for Cooper rules engine, and just simplifying a lot of the things that are there with EdgeX today like the deployment orchestration with new compose files.
Swapnil Bhartiya: And I was saying the last thing when we talked was almost a year ago during the hackathon where we also saw a lot of, how organizations were trying to use regex with a lot of other projects. So I think it’s almost one year, or more than that. What kind of new use cases have you seen so far? And I’m talking about use cases where it’s been used in production, not, you know, evaluation, or experimentations.
Jim White: That’s really where we’ve played a lot of attention here over the last six months, even for a year. This past summer, we sponsored a set of what we call adopter series presentations. So these are people out in the field actually using EdgeX and putting EdgeX to use in their product sets. And we’ve seen a number of those. And it really helped to grow our community. So some of the major ones that came through with some of the adapter series presentations include ThunderSoft, which is a company out of China big SI. They’re using it as part of what they call Thunder Cloud; so an actual IoT edge deployment capability. We’ve got Accenture, which has made it part of their AIP plus application. So another big SI focus globally, and using EdgeX really to get data into their big data analytics. And then you’ve got companies like HP. They have come out with a product they call engage, which is really using EdgeX at retail point of sale locations for things like loss prevention. So we’ve seen an explosion in use, we’ve seen people who are adopting and actually coming forth and saying, “Hey, here’s how we’re using it”. They’re also telling us what’s been great for our community as well as these are some of the features we’d like to see going forward. So they’re really helping us to shape the project for the future releases.
Swapnil Bhartiya: Thanks for sharing those use cases. The fact is, Edge is becoming really big. We recently did a whole in our series of prediction videos, and everybody is talking about Edge. But we are not talking just Edge as you know, Edge devices. You know, they’re also talking about Edge data center and all those open source technologies pretty Kubernetes. Also, 5G is playing a critical role. So what role do you think is edX lb your regex projects good to play in this space?
Jim White: Yes, Swapnil EdgeX has always been and will continue to be Edge middleware. And I tell people to try to simplify it as much as possible meaning we really sit between the things that IoT sensors and devices, and whatever lies at the next level up at your enterprise or your cloud layer. So while all these technologies are coming together to try to participate in the Edge, what most organizations and what’s most of those products you just mentioned, need is something that really gets down to that lowest layer. And so that’s what EdgeX has and will always continue to do is provide the connectivity tissue between the actual things the sensors and devices. And whatever you’re trying to get that data to. So we’re excited about it. We think things like 5G are going to see an explosion, obviously and the way that people try to incorporate Edge into their products and solutions. And that just makes subjects all the more relevant.
Swapnil Bhartiya: One more thing is earlier, during the hackathon, what we saw was that there was no kind of standalone project. No. In today’s world, when we build either a stack or application and fracture, we use a mix of different open source projects not to build different pieces, you know, so what projects are there, either within the Linux Foundation or outside the Linux Foundation that EdgeX relies on our collaborate with so that users, as you also mentioned, you know, can leverage all these technologies. We got them to basically care about what they want to deliver to their users?
Jim White: Yeah, exactly right. And we’ve really built objects like that from the beginning, which is to say, we’ve built it as a micro service flexible architecture. We’ve never believed that an object is going to provide everything you need out of the box. We try to provide quite a bit in what we call a reference implementation. But we both leverage a lot of other open source capability. And we connect to a lot of other open source technology. A case in point is, we’ve always had a rules engine as part of objects to provide local analytics. But that’s not a rules engine we created. We initially came out with drools rules engine connectivity, and now we’ve switched over to use mq Xs Cooper, which is an open source rules engine being developed in China and one that’s really easy to use, because it’s SQL based. At the same time, we were always providing connectivity to things like MTT brokers as well as trying to provide connectivity to other cloud and back end systems. At the same time, we’re incorporating what we see as best of breed solutions into our product, because we don’t want to have to create it all. So for example, the database layer, we use Open Source Redis as a means to provide for persistence at the Edge. And we’re using things like Redis Streams, where people are wanting streaming capability now with Ajax. So we both use a lot of open source technology, and then also incorporate the ability to connect to and incorporate irreparably a lot of technology as well. Collaboration is key to us. I mean, we came in as one of the founding members of elephants. And so we believe in a strong community. And I think, you know, it has been alluded to both in the blog posts and what have you, we put some new features in there, this time to provide integration points to fledge project, which is a sister project that is also serving a bit of thin IoT edge middleware. But in addition, we continue to collaborate with open horizon which we incubated underneath EdgeX. We are looking at a new project coming up in 2021, to work with sto open horizon, kind of providing a full stack Edge solution along with EdgeX, obviously, as part of that, and actually coming up with an occasional blueprint to help define it all. So collaboration is something we do both internal to lF and outside of lF. And one of the things that we kind of as an open source effort just kind of take as part of our tenets of behavior in terms of how do you make Ajax more relevant by making it more interoperable, and integrated with other things out there?
Swapnil Bhartiya : We are almost at the end of this year. And this is, you know, semi enrollees. What are the things that are there in the pipeline? As it is an open source project, anybody can go and check, but still talk about the things that you are looking forward to incorporating in the next releases.
Jim White: Yeah, we’ve got a really awesome group. And they’ve put together a fantastic roadmap. So for 2021 will have two releases. The first release will come out in the spring, sometime around April or may slip a little bit into May, because this is going to be a big release. This is going to be EdgeX 2.0. So as I mentioned, while we’ve started working on a new API set will actually finish that API set and make it the API set of views starting with this next release. And that’s really going to eliminate a lot of technical debt, and provide for a lot of new features, including more messaging and objects going forward. Beyond that, we’ve got a lot of other features that we’re looking at; things like the addition of metrics, reporting from objects, so you can get a sense of how things are running and what’s operating (well, what’s not operating). We’re actually putting a lot of performance and scalability testing in place to really be able to start to offer users guidance about how much and how fast objects can move things. So that’s gonna be coming up in our next release, the spring release. Beyond that, the release after that will be in the fall. We think that’s going to be a relatively small dot release. And really, it’s what we’re trying to call a stability release. And the idea is once we come out with 2.0, you want to release where things are a little bit tamped down a little bit, you don’t have quite as many features, so that you can really say that this release is you know, whether it be object 2.1 or 2.0. point one or something of that nature will be released that people really start to say, this is where long term support is going to be offered by objects. In fact, that’s our hope with the release in the fall. We’ll have our very first long term support release.
Swapnil Bhartiya: Do you have different releases like other open source projects where there is one stable release that people can use in production and then there is an unstable release for testing new features that gets frequent updates so that while you can move at your own pace, without disrupting the way users are using it?
Jim White: Yeah. And that’s exactly what we do, something similar to a lot of open source communities. We’ve really taken their lead. We’ve got some great folks from Canonical that have really helped, describe it and make it understandable to us and the rest of the community about how to do that effectively. So yes, you’re exactly right. We haven’t had a long term release, yet long term support release yet for that very reason. You know, we really wanted something that was stable that could last out there, by our own rules last at least two years. And that’s why inject 2.0 is going to be important, and then the LTS release right after that. Beyond that, we do release every six months. And so for those who are looking for the latest features, those are looking to really stay on the cusp of what Edge is doing. Every six months, here comes a new release, which will really help you kind of get the new features you’re looking for and add on to it as you see fit. And then we’ll incorporate that in the LTS over time.
Swapnil Bhartiya: One more thing that I wanted to ask you was if security is becoming really serious, especially in the use cases that we play with, because sometimes things are outside of your corporate firewall. How important is security for EdgeX Foundry? And then number two is that wherever you sit in people just tack as you explained earlier? Do you really need to worry about security at that stage? Or the other pieces before and after objects foundry are taking care of it?
Jim White: Yeah, let me address both questions with the answer of Yes. Both are important, right, both having a security index is very important. But then also you’re going to need to tie in a more cohesive strategic security strategy around your whole deployment and use case in EdgeX. We’ve got a fantastic security working group led by Collin Hutchison, who’s a member of Kong, so very familiar with, you know, things not only at theEdge, but also the enterprise and how to secure things like API at the entire ecosystem. So we provide means to secure the API’s and objects and we also provide a secure store. Those are the two features that we really, you know, put at the heart of our product, but those would be at the heart of many products like this. Beyond that, you have to start to think about a lot of other security concerns, some of which we try to help with within objects. Others, you’re going to probably have to start to think about that from your strategy. I know waking up and ending up on the front cover of The Wall Street Journal is something that someone has been hacked or an organization that is putting out nonsecure product is what keeps me up most at night. So we’re trying to do our best. But this is a difficult problem just being because the edge is such a different animal compared to what most people are used to. And so that is some of the feature set that we’re both incorporating in our products today. But it’s certainly part of a larger roadmap as we go forward is how do we make the edge as secure as possible, at least the data that flows through EdgeX, so stay tuned and more will happen in that space for sure is our communities is really looking at that, too.
Swapnil Bhartiya: Thank you so much for your time today. I look forward to talking to you again soon. Thank you.
Jim White: Thank you. Yeah, it’s good to see ya, at least on video Swapnil hopefully we’ll see each other in person soon.