Emerging Technology and Open Source Software at Red Hat | Parul Singh

Luke: In this episode of In the Open with Luke and Joe, we bring you a conversation with Parul Singh, a senior software engineer on Red Hat's emerging technology team. We will be discussing the open- source projects VolSync, MicroShift, and The ChRIS Project, as well as her work with OpenShift's quantum operators. Before we welcome our guest, let's say hello to our co- host, Joe Sepi.

Joe Sepi: Hey, Luke. How are you, my friend? How's the weather over there?

Luke: I'm doing well, Joe. The weather, it's getting cool. I feel like autumn... I can feel it in the air, although we've had a lot of rain in the Northeast. You may have seen flooding. Serious flooding affected a lot of people in New York, Pennsylvania, New Jersey.

Joe Sepi: Yeah, the flooding was incredible. I think that we're both fortunate that we didn't experience it to that level. Certainly downpours. It was gushing over here, but no extreme flooding like I was seeing the videos in the city and stuff. It's really wild. But yeah, beautiful the day after, and today it is gorgeous out. It's fall weather. I've got a lot of light jackets that I'm eager to pull out of the closet. So, that's nice. That's real nice.

Luke: We've mentioned that several times on this show, and I feel like we've been teasing the audience. I think we need to do Fashion Week, where we see some of your fall jackets.

Joe Sepi: Oh, I got this one. I can't wait to show you. And I got a funny story about how I left it in San Francisco at a conference, and one of my colleagues when I worked with New York Times actually went around to all the places we went, and he found it for me and brought it back. We'll save that for another episode.

Luke: True friendship.

Joe Sepi: So true.

Luke: Without further ado, let's welcome our guest, Parul.

Joe Sepi: Hi.

Parul Singh: Hello.

Joe Sepi: How are you?

Parul Singh: I'm going well. How are you guys?

Joe Sepi: Good. Good. You're not far from us too, right? You're over in Boston. How's the weather up there?

Parul Singh: Oh, don't ask. It's been raining Past few days, but today I feel like it's a little sunny. So, my plans will like the weather today.

Joe Sepi: Yeah. Yeah, and I'm hoping for good weather this weekend as well. That'd be nice. So why don't we start with it, Parul? If you don't mind introducing yourself, we'll go from there.

Parul Singh: Okay. So as you know, I am Parul Singh, and I work as a senior software engineer in the office of CTO, and I work for the emerging technology group. What our group does is, we try to think a year or half in advance, and see what are the emerging tech and how they can be identified and aligned with the Red Hat technology. So my job is to go, and find, and research the emerging technology trend, and then make cloud- native prototypes, and tie that back to find opportunities that can be aligned with Red Hat's strategy.

Joe Sepi: Excellent. There, I find emerging tech really interesting and a really fun place to be doing that sort of work, development work, and research. How long have you been there, and what are you excited about, and what are you working on these days?

Parul Singh: I am relatively new to this group. So I joined a year- and- a- half, and I've been working at Red Hat for almost just two years. So a year- and- a- half I was brought into emerging technology team, and I've been working on multiple projects. But the reason that I was brought into this team was to develop the IBM and OpenShift quantum operators. For that, we have been using OpenShift to interact with IBM Quantum backend devices. And the idea is how you can bring quantum on cloud, and democratize the access to quantum computers using open- source API.

Joe Sepi: That's fascinating. Luke and I have gone and seen the quantum computers at the IBM facilities, and they're amazing, and there are these huge rooms that need to be stabilized physically, and everything needs to be kept at a certain temperature. So, how does this work in terms of the quantum work that you are doing with OpenShift? That's a big, crazy, super- special computer, but you're able to do it with OpenShift. How does that work?

Parul Singh: So the idea behind is, not all task or not all jobs are best for quantum computers, and they are also a limitation of classical computer. So our vision was, " How can we get the best of the both world, and how can you offload specific tasks that quantum computers are best at solving to quantum computer while have the rest of the workflows on the classical computer?" So that's why we developed these operators that can offload your quantum section of the job to the quantum computers, while have the rest of the workflows running on OpenShift. I've been working with the IBM Quantum team. Ismail, and rest of the people were from Spain. So we developed these bunch of operators, and one of the operator is specifically just to help you get started on developing the circuits. So imagine, if you are a quantum physicist or you are quantum circuit writer, you are not a conventional software developer. And getting started installing the development environment, installing the libraries, understanding what are the dependencies, this is not what you're required to do, and it could be very overwhelming. So we develop this operator that, just with a click of button, you can have the entire development environment set. And you can just start developing the circuits, because that's what your job is.

Joe Sepi: Yeah. This is fascinating to me, because again, I think about quantum as this crazy, wild computer, and I feel like it has felt so futuristic for a while. But if not paying attention, it's slowly becoming normalized. I think about that in the sense of OpenShift, but also I know Qiskit... If that's how you say it. Q-wiskit? Qiskit?

Parul Singh: Qiskit.

Joe Sepi: Yeah.

Parul Singh: Everybody has their own pronunciation. But the creators of Qiskit call it Qiskit, so I call it Qiskit.

Joe Sepi: Yeah. Yeah. That's an open- source framework for working with the quantum machines. So my point is that the computer's wild and feels futuristic, but all of the things around it are starting to become the things that we're used to doing. Kubernetes, and OpenShift, and open- source frameworks to access it. And it just feels like it's becoming more normal.

Parul Singh: Yeah. It is an ever- evolving technology, believe me. The work or the APIs that I was working for a year, they need to be changed right now. So it's a very fast- moving sector, and we are just trying to scramble and keep ourself up- to- date. It is a time where people have started thinking quantum, and they are thinking how they can apply quantum to optimization. Initially, people started with thinking that" Oh, if I apply quantum computers, I am just going to change the computational speed exponentially." But as of now, if you see how the market is evolving, even getting a 5% optimization is a huge deal. Imagine if you are fleet management service, and you can optimize your route and your network even by 20% or 10%. It's a huge deal.

Luke: I dropped a link down below if anyone's interested in seeing the quantum development roadmap from IBM. And as we're mentioning, we're in this very interesting time. I liken it to the 1960s at NASA, where all of this stuff is happening and going from science experiments to real- deal, real- world stuff as we speak. So if you want to check out that development roadmap, that's a great place to look at. And as you mentioned, it's very interesting. I went to a financial conference right before the pandemic hit, and while they may not be ready to use it today, all of these banks, all of these financial organizations are working on it now. Because they need to figure out the algorithms, they need to figure out the workflows, they need to get ready for when it is real. You got to be competitive. So, very exciting space. Please, I put this in the chat, and I will put it in the show notes. One other note for everyone who's listening. If you have any questions for Parul, please drop them in the chat on whatever platform you're listening and we'll get to them towards the end of the show.

Joe Sepi: Yeah. I don't think we plan to dig into this too much, but I'm curious if you could share... Because I think quantum is really interesting and exciting. Can you share different technologies or fields that people might be thinking about quantum already, and trying to work on that? I know Luke had mentioned finance, and you had mentioned logistics, I guess you'd call it. What sort of things are people thinking about in the quantum space?

Parul Singh: There are many, of course, but it has been applied at also things like chemistry. Whenever I'm giving a quantum talk, this is my favorite example, is as of now, you cannot represent a single molecule of caffeine with classical computers that you have. Even the best classical supercomputer that you have cannot represent a molecule of caffeine.

Joe Sepi: That's wild.

Parul Singh: And if you want to go ahead and build that, it would consume almost 1/ 10 of the atoms on the Earth. But you couldn't do that with just... I don't remember the exact qubits. But it is manageable using a quantum computer. If you have a quantum computer that scale of 64 qubits, you can represent a molecule of caffeine. So it's so interesting, and the example, why I gave this is because the application of quantum computer in chemistry. You try to find out what is the best molecular structure of medicines or fertilizers, and you don't even have to actually make them and test them. You can just build an algorithm that come up with different molecular configuration, and you can decide, and you can learn from those medical configuration which is the best or the best- suited for the job it seeks to. Also in AI, think about automotive cars. If you're able to make these instant optimizations on routing and how to navigate through the road system, even though you could also apply edge. So there's so many overlapping scenarios, but you can take it to AI, you can take it to chemistry, you can take it to physics. It has applications almost everywhere.

Joe Sepi: Yeah. Isn't the traveling salesman, isn't that kind of another classic example that they talk about quantum, or am I confusing that to you?

Parul Singh: Yeah. Any optimization problem. If you want to find the best path, the best route, using quantum computers you can do it very easily as compared to classical computers. And faster.

Joe Sepi: Yeah.

Luke: It's so interesting. Yeah. Let's come back to quantum. I want to make sure we get in all of our topics today. Take us back. You started working on the ChRIS Project, and this is the beginning of your open- source journey, and it brought you into so many other projects. So what is the ChRIS project, and how did you get involved?

Parul Singh: So in the final year of my grad university year I was connected with Mass Open Cloud, which is an open cloud in the Boston area, and it's made of a collaboration with universities. Harvard, MIT, inaudible, my university, Northeastern, NBU. I did this graduate research assistantship in MOC, and from there I was introduced to the ChRIS Project, and I've been intern on the ChRIS Project. So those of you who don't know, ChRIS Project is a collaboration between Red Hat, Boston University, and Boston Children Hospital, and we are developing medical imaging software that could make clinical results more relevant. By that, to say that right now, if you have to analyze thousands of MRI, it takes a lot of time. And let's say that you had an infant who has a very rare brain disease. It would take you a lot of time to even find out, because you don't have as many samples, or the doctor has not himself or themselves seen such scenarios. But if you reduce that time from days to minutes and hours, it is more relevant. You can start attacking the disease. You can find a medical path to take for the prognosis. So that's what I've been working on, the ChRIS Project, and that opened the door for my journey to open source. And then I was hired at Red Hat, and I continued to work on the ChRIS Project, which belongs to the research team at Red Hat. And I was there for a year, and later, after a year, I moved to the emerging tech team.

Luke: So interesting. It's interesting because it's open source, but also that it has this tech- for- good, but real- world application as well. It's not some sort of theoretical thing. It's doing good immediately. Open source and a tech- for- good. I love it.

Parul Singh: I think that's the power of open source. Sometimes you can see the impact happening so close to you, and it does not need to take two or three decades for it to actually make an impact. You can see that it's just happening right there, and you don't need to be... For example, if somebody had to work on this, either they had to be an employee of Boston Children Hospital or they had to be an employee of Red Hat. But a student like me got an opportunity just because it's open source, and anybody can go and contribute into it. So, that's the power of open source.

Joe Sepi: Yeah. I think too, one of the other kind of smaller aspects of open source that we talk about sometimes here is, it sounds like you got a job through open source. Which is another great benefit of doing open source. People see the work that you're doing, they interact with you, they find you to be professional, easy- to- work- with, but also doing great work, and jobs come out of that. I've seen that happen many times. It's great.

Parul Singh: Yeah. Whatever work you do, it's out in the open. Anybody can go and analyze that. It's not proprietary software, that you don't see what you're making.

Joe Sepi: Yeah, exactly.

Parul Singh: It's out in the open.

Joe Sepi: Yeah. In the open. Well- played. So yeah, this is interesting. Is this something you still work on, or do you know what the status of the ChRIS Project is currently?

Parul Singh: So, my interaction or the time I spent has been reducing significantly. I started with being an actual contributor, then I was managing that project, and now I just try to connect the people with the right team if they have any ChRIS- Project- related questions.

Joe Sepi: Very cool. Yeah. That's really fascinating. You've got the link up there. I encourage people to check this out, because it's really quite interesting, and a great, practical application.

Luke: I only briefly looked over it, but what I liked about it too was, it seems like the way it's set up is also a very modern or contemporary computing structure. It's using containers, it's using compute clusters. So I feel like, besides doing good, there's a lot to learn from just how it's executed.

Parul Singh: So we do use OpenShift in ChRIS Project as well. Again, the idea is how you can scale this system to immense power so that you can do the computation really fast. And my contribution to this project was to introduce a scale, and the portion I developed was how you can run it on OpenShift, and you can do horizontal scaling from one pod to 10,000 pods as needed. So, it has all the new technology. It uses containers, it uses OpenShift, which is the heart of all the cluster.

Joe Sepi: Mm- hmm. Very cool.

Luke: What comes to mind too, Joe, hearing this, is in our last episode we had Ted Tanner and Adam Orentlicher on from Watson Health. One of the projects they were talking about was LinuxForHealth, and about how edge is such a big component of that because of huge data sets, because of security, because of network considerations. And I hear about this project, I hear that it's using OpenShift. And then talking about some of the projects we're about to talk about, I could see MicroShift being on LinuxForHealth. It weird. There's a lot of connections to be made here.

Joe Sepi: Yeah. Yeah, and James Snell before that one as well. Yeah. We've got a thread going here that I don't think we really even planned, but it's fascinating.

Luke: I wanted to ask you about the VolSync. I think that's a good next step in our journey of exploring this space. So give us the high- level overview, and tell us what aspect of it you're working on.

Parul Singh: The VolSync project started with how you can do data mobility on cluster. It is very easy to move applications, but if you have to move stateful applications that have data and that have states, it's very difficult. Not just difficult. It also depends on the storage system you use, and it has to be that the primary site and the secondary site has to be the same storage. Right now, how people do it is, they rely on the... Whatever the storage system they're using, they expect that it has a data mobility feature that can help you to move data. But you could just imagine how much reliance not just on the storage system for storing the data, but also managing it, backup, and disaster recovery is. So we came up with this operator, the VolSync operator. Basically what it does is, it provides a storage- system- independent mechanism for data mobility. And the use case you can think of is not just simple disaster recovery or backup. You can also do data replication. You can manage data in a multi- cluster environment using simple GitOps operation. So, that was the motivation of bringing VolSync. How to make data mobility in clusters very easy. And the specific case that I was working on was how you can replicate data in a one- to- many high- fan- out scenario, where you have a primary cluster from where you are distributing or replicating data to multiple edge devices. That's a specific use case that I've been working on, and quite interesting. I gave a talk about it yesterday. Go and check it out. Search DEF CON VolSync and you will find that presentation. So the use case is, how do you do it using the underlying Kubernetes CSI components to do data mobility, and not actually relying on the storage system itself?

Luke: So interesting, and so exciting that you presented at DEF CON. What a cool event to have on the resume.

Joe Sepi: Yeah, really cool. So I'm reminded of, and I want to ask if this would be applicable here. I saw talk a while ago. This woman was working in Africa. And now my brain is not working very well, so I can't remember what the disease was. The pandemic was there. But she was trying to do tracing, and a lot of people were, in an area that they didn't have a lot of connectivity. So they would do it locally, and then when they got online somewhere, then things would sync up. And I guess I kind of wonder, does VolSync do that sort of stuff as well? Offline, and syncing, and replicating, but also being able to deal with limited connectivity and such?

Parul Singh: So for VolSync, if you're using VolSync, you would need to have at least a network connectivity when you are replicating. But it's so cool, because it's schedule- based. So, let's say that you don't have access to network all the times. But during the time of the week or time of the day when you have access to a network, you can schedule your operations to run at that particular moment. You can set the schedule in the operator itself, and whenever the schedule is on, the operator are replicating the data. So how it happens is, you have a primary cluster, and we imagine that the primary cluster could have internet connectivity all the time. If not, again, on a schedule. And the primary cluster puts all the data onto an intermediate storage unit. That could be an S3 bucket in an object store. And all the edge devices that you have, that also runs on a schedule. And whenever the schedule is on and it has internet connectivity, it will just download the data from the intermediate storage unit, and it will provision the underlying volume on those devices to get the data that was replicated recently. So, you don't need internet connection all the time. But since it is schedule- based, and if you ensure that, at the time the schedule is on, there is network connectivity, it will do the job. Actually, that was the idea for the use case that we were working on, is, it is for edge devices. Imagine drones that are set out to get pictures, and they have limited space also. So, they can't store a lot of data on those drones. But if they're running on a schedule, they go and collect the data, they go to the far North Pole and South Pole. They collect the data, and whenever they are near the observatory or laboratory they have internet connection, and they would just copy, or replicate the data, or download the data, or upload the data. So this example, I guess maybe it was Ebola, because that is what inaudible.

Joe Sepi: Yeah. Yeah. Yeah. Yep.

Parul Singh: Yeah. So, it is for use cases like this. For edge devices that have limited network, that have limited storage, and they can use VolSync to replicate the data.

Joe Sepi: Yeah. That's fascinating. I know tech can't solve all of the world's problems, and it would be a fool to try to do that. But I think it's really interesting to try and think of ways that we can apply technology to be helpful, and solve some of these problems or parts of the problems, and tech- for- good.

Luke: I just thought of something really quick. I think we may be taking for granted that we understand what operators are and how they plug into OpenShift. So for our listeners who are like, " What is this operator? It sounds like magic." Maybe Parul, you could just give us that high- level overview.

Parul Singh: Sure. Think in the old time, when you had to manage an application, you are responsible to seeing, " Is it updated? Is there a security patch that needs to be applied?" If it's down, you have to restart it. So in a very simple term, operator takes all these tasks from a human, and it is like a binary that would be managing your application. So not just running the application, but also updating the application, applying any security patches, or any vulnerabilities that stay in the software. And when I'm talking about operators, yeah, I'm specifically talking about Kubernetes and OpenShift in mind, like when you're developing operators for Kubernetes. The VolSync operator is Kubernetes operator.

Luke: Excellent. Sometimes it's so easy to forget that there's things we need to unpack and explore a little bit.

Joe Sepi: Yeah. I thought that a few times too. We should take a moment and explain what operators are.

Luke: I was also going to mention a use case for this that came to mind. I recently got involved with some agriculture projects, and there's this... I think he's actually listening. My colleague, Joe Pearson, is part of the Open Horizon open- source project, which they use in factories and all sorts of things. But that software is now being applied for agriculture, and this exact kind of use case comes up, exactly like you're saying. A field or a ranch, it's a huge space. You're not going to have connectivity. But back at that barn or that ranch house, there will be. Right? So maybe the tractors, or the drones, or whatever's traveling and traversing that property, when it goes back, it can update that data, update its models. And then go offline, do what it needs to do. Collecting new data, applying those models. And then when it comes back to that point of connectivity... And I think when we start talking about the MicroShift, I think we have some other really neat edge cases that we're going to bring up around the CubeSat. Which is, to me the ultimate edge case is the outer space. Right?

Parul Singh: Yeah. So I would say that MicroShift is a bare- minimal Kubernetes cluster. It's like a flavor of Kubernetes, and you can think of VolSync is an operator that would be running... Just for people who don't have very much tech background, the difference is, VolSync would be an application that would be running on MicroShift, if at all, and MicroShift is a bare- minimal flavor of Kubernetes. You strip off all the things that you don't need, and you just have made it... It can work in adverse network connection. It can work in adverse storage and CPU memory. So it's very minimal flavor of Kubernetes that is designed for far- edge devices. I saw the tweet from Naeem yesterday, and he mentioned MicroShift. So, that was pretty cool. That just made my day, and was so fortunate that I'm working on MicroShift

Luke: If folks are interested, Naeem is the space CTO at IBM, and we've had him on the show before. He's involved in lots of cool projects, and one of them is a CubeSat. But very interesting project. We'll put it in the show notes as well. But it's interesting to hear that yeah, MicroShift is going to be running on the CubeSat in outer space. We should get into MicroShift, but maybe... We're about halfway through our time. Maybe we should do a little break here, let folks know about what's going on with IBM developer and podcasts. And then when we come back from that break, we're going to talk about MicroShift.

Joe Sepi: Sounds good.

Luke: So, thank you for tuning in. Again, if you have any questions, please drop them in the chat. We'll get to some of them towards the end of the show. Also, feel free to tweet at us. We're happy to engage. Yeah. You can always find our latest show at the top of ibm. biz/intheopen, as well as all the archived shows below. There's also a link on that page to our podcast page, where you can find the audio version and links to your platforms of choice. Yeah, and there's also a direct link to those podcasts. So if you don't feel like watching the stream or the replay of the stream, and you want to catch it as a podcast. I also clean up all the ums, uhs, and buts on there. So it's a more concise, tidier version. I sound much more eloquent. Joe's always eloquent. Our guests are always eloquent. It's I'm cleaning up. I would also mention developer. ibm. com, the parent page for all of these podcasts. There's all kinds of great content on there. Blogs, videos, all kinds of enterprise and open- source tech on there I would recommend checking out. Let's shift back to our topic at- hand MicroShift. So yeah, give us the rundown about the high level of what's it used for all over the place, and then what part are you working on?

Parul Singh: So, first thing. When I work in emerging tech, it's very important to know that we don't work on any products. We are working on pre- productizing team. Everything that we are doing is in a pre- productizing stage. It's not actually released for commercial. So MicroShift is an experimental flavor of OpenShift and Kubernetes, and it is optimized for edge devices. By that, the target use cases are your standalone edge systems that uses Linux to its clusters that runs on OpenShift and Kubernetes. So MicroShift is targeting the bare- minimum things that you can have in a Kubernetes cluster to run your workflows, and it relies on the underlying operating system, security configuration, rollback, and updates mechanism. And MicroShift itself, it run just as a deployment on the operating system. So what you can do is, you can just image a device, and you can take that image, and go onsite, and plug it to a network. And boom, your applications and your workloads would be on those devices. So, it's just a very simple plug- and- play that needs bare- minimum networking connections. It can also work on disconnected devices, because if you're able to carry an image and just plug it into the edge site, it can be run as a disconnected service as well. As long as you ensure that all the images that you have, all the container images that you have, are available on those systems so that it can just pull those images from the image registry. Yeah. So that's the nutshell about MicroShift, and what I am working on is... It's a very young project, so we work on everything as whoever gets to tackle what they're doing. The part of work that I'm doing is just how you can inaudible the code to have an internal service manager to start all the Kubernetes and OpenShift services. So, nothing specific. Everybody's working on everything on MicroShift.

Joe Sepi: That's really cool. I wonder if we should step back just a moment, and we talk about edge computing. And I think we've touched on it a little bit here, but what are some good examples of edge computing? And why, for our listeners, do you need to adapt for those environments or those computers, those edge computers?

Parul Singh: You mean to say the kind of use cases?

Joe Sepi: Yeah. Yeah, and they're typically smaller computers. Raspberry Pis, things like that, so you have to build these things to run on these different devices.

Parul Singh: Right. So we also have, for the CubeSat, we are creating the multi- arc image of MicroShift so that it could run not just on x86, but it can also run on ARM, on the underlying device that is compatible with the whole CUBSET ecosystem technology. So, it runs on multiple architecture. It's not just really made for x86. It works on Linux. It works on Windows. It works on MacOS. So, it is optimized for all the different OS as well.

Joe Sepi: Very cool.

Luke: Very cool.

Parul Singh: Yeah. And you have to think about, what are the minimal things that you can have on an edge system? Because an edge device is not a full- fledged cluster that you can have, like a server. So just think, wherever you cannot afford to have an entire server system running, you can run MicroShift on the bare= minimum resources that you can have.

Joe Sepi: Interesting. You got a link there, Luke? I saw you pull something up.

Luke: Yeah. I was going to say, I totally got the name wrong. It's ENDURANCE, is the name of the CubeSat.

Joe Sepi: Oh, cool.

Luke: But yeah, this is the CubeSat project, and there's something where you can... I think it's still open, where you can submit and get your name on the CubeSat, which is a-

Parul Singh: Yeah.

Luke: And it uses a quantum computer to generate the UID or something. It's very interesting.

Parul Singh: Yeah. I am trying to find that tweet. I'm just not able to go at this moment, but I will share it with you once I have it.

Luke: I think, to answer your question too, Joe, something that you were getting at there. It's the unique edge cases of edge. I think an example too would be self-driving cars.

Parul Singh: That car is a great example. Even if you had seen, I think Naeem was in inaudible, in the open inaudible earlier, and he was also talking about when you are running these experiments in the space. And it's at that time when, again, you have very limited resources over there. So it's anything that is disconnected network, limited resources, limited memory, limited CPU. That's, we call it far- edge devices, and MicroShift is suited for those kind of deployments.

Joe Sepi: Are we thinking too... And these are, like you said, mission- critical kind of things and whatnot. But also, my toaster oven and all sorts of everyday items. These are considered edge devices as well, and will become smarter and more capable to do all sorts of things that maybe we don't want them to do. I don't know.

Luke: Well, it's funny you bring that up, Joe, because they don't seem like they're mission- critical. But something else I know we wanted to talk about today was sustainability, and I think there is a big connection here. I know some air conditioners already have network connections so the power company could turn them off in case of a power crisis, because that's what could tip the scales. It seems trivial, but something like our refrigerators, and our toaster ovens, and these things at- scale, altogether. Actually, I think there is a use case there for being able to collect data from them and potentially being able to communicate during times of crisis.

Joe Sepi: Yeah. No, that makes a lot of sense. Fascinating. That's perhaps another good segue to talking about sustainability and the environment. I know that's something that you're very interested in, Parul. You want to share a little bit about that?

Parul Singh: Personally, I like to live as sustainable and as green as possible. I have not consumed single- use plastic for a year, and that's something I'm very proud of. So one of my hobbies is, I've picked up woodworking. What I do is, in my current house, except for my couch, all the furniture has been built by myself, and I don't use any new material for that. People leave a lot of stuff on the curbside. So whenever I find something that can be wiped and that can be upcycled, I just get that. I strip it off. The paint, the coating, the stain, everything. It's a grueling process. It's very tough to live a sustainable life. It's not easy, but I made that oath that I would not be buying anything new. As much as I can avoid, I don't buy anything new. I try to refurbish, or reuse, or upcycle anything that's available. So, that's the sustainability life I'm trying to live.

Joe Sepi: That's amazing. You've been doing that for a year, you said?

Parul Singh: Yeah. In the pandemic I picked that up, because I had nothing else to do. You couldn't go out. So in the basement, I just... I've been collecting these scrap pieces of wood for a long time, having in mind that something I'm going to do out of it. In the pandemic I just got all my collected junk, and then made something out of it.

Luke: I knew we wanted to talk about sustainability, but I had no idea that this was the rabbit hole we were going to encounter. Just to reflect here, it's so interesting. A lot of the folks you meet within the tech and development industry, you respect what they're working on at work. Very interesting. And then you scratch the surface and you find that, " Oh, wow. There's so many other interesting hobbies and avocations around a folks in development." So happy to hear you share that. Similar, I have all kinds of tools, and I'm renovating a house that was a shell, and I've been building it around myself for the last year. It's been harrowing.

Parul Singh: Yeah, but it's really cool to try to savage old stuffs and not to pollute more. Because if you think, everything that you buy is eventually going to end in landfill, and it is such a daunting thought. One day I was just sitting like, " Oh, no. I cannot do this." And you wouldn't believe, when I was in my early 20s, I used to buy a lot of clothes. I was falling fast- fashion. But then there was a time when I realized everything that I buy is going landfill, and they're going to live on Earth for more number of years than I'm going to, and that was so terrifying.

Joe Sepi: Yeah, it's so true. And I think even if you aren't fully onboard with the sustainable, not buying anything new, I think there's still a lot of value in just finding interesting things and refreshing them. We found an old dresser that we thought was really interesting, and you strip it down, and put a cool color on it and some hardware, and now you've got this really interesting piece of furniture. I'm curious. I'm trying to imagine. Do you have a garage out there, and a big woodworking setup, and ...

Parul Singh: No. I am like an edge device. I work in adverse conditions. I live in a condo, and I go in my basement and I chop wood. It is really loud, so I'm just waiting to get a notice. " Oh, you are a very loud neighbor and tenant. Get out of the building."

Joe Sepi: Yeah. But yeah, that's funny. I have an apartment here too. So whenever I'm doing some work, when I'm at the apartment, I'm always like, " Oh, the neighbors are going to hear me."

Luke: Are they complaining about your guitars, Joe? Is that happening yet?

Joe Sepi: No, because I have some amps here, but I don't plug them in. I just plug them into the computer and put my headphones on. But I keep daydreaming about setting up a drum kit, but I don't think my neighbors are going to like that. We'll see. We'll see.

Luke: So this question, it's not directly related, but it's such a perfect question for a Red Hatter. I thought we would bring this in real quick here. Shaw asks, " How can a company earn money from open- source software?" Red Hat's the best company in the world to ask this question.

Parul Singh: Okay. So I will tell you what I was told in my orientation program is how you earn money from open- source software. Think of open- source software as water running in a stream, in a mountain stream. That is free, right? It has all the minerals, all the salts that you want. And it's spring water, so crave after. So if you want to drink spring water, what are you going to do? You can hike up the mountain, you can bring your tumbler, you can go to the mountain stream, and you can drink it. Or you can get a company that packages that water for you, put it into a bottle, and give it to you. So, open source does not mean free. Open source means that it is available for everyone to see. Everybody can have the code. And if you are willing to put that much energy into managing it, ensuring that it is secure, it has always updated, then you can use it for free. You can just clone the GitHub, and you can just start building your own executable. But if you don't have that much money, if you don't... Sorry. That much time and resources, then you just take it from someone who will do that for you. So, that's how open- source company makes money.

Joe Sepi: That's a great analogy. I'm not sure what the term is, but that's a great metaphor. I'm going to use that. I remember seeing, when I lived up around the Saratoga Springs, there were a few spots where people could go and fill up their plastic jugs. They didn't have to hike up the mountain, but even still. I remember we did that a few times, but most of the time we just got our own water. So, that's an example I can really relate to.

Luke: That is a great example, and I like too how it illustrates when and where you would want to pay for it. If you live on the mountain, okay, great. Just get the water right there. But how many of us live on the mountain next to the spring? And you have stuff to do. I'm trying to make breakfast. I need that water. I can't hike 10 miles of the mountain and back just to do that. So it's interesting to find that breaking point of, "Oh, this is why I want to use a product based on a project, because it actually does provide that value for where I'm at and what my resources..." And like you said, time is the biggest resource. How quickly can I get to value for my customers and my industry?

Parul Singh: Yeah. And still, it's open source, so if there is any ... For example, if you are at Red Hat, we have lots of upstream projects. And from that, we make a downstream. So all those innovations and creativity, it still goes in the upstream project. And if you're buying well from us, you are ensuring those innovation and creativity are still there. But you're also ensuring that it is secure, it is reliable, there would be somebody to help you if something goes down. So that, again, is one of the why company follow or would like to have an open- source ecosystem.

Joe Sepi: So, very interesting. I'm curious, if we kind of shift gears a little bit here, you working in emerging technologies. What are you excited about? Is there upcoming work that you're expecting to do, or what's got you excited about the work that you're doing?

Parul Singh: The things that really excites me is the innovation, because we are thinking a year in advance. And it's like the actual team or the products, it's not realized, it's not released. So, I like to say that we are the trendsetter. We go and see what is going to happen, what the market is going to evolve in. And we do a lot of prototyping and analyzing the whole sector, and then coming up with solution, and then working on the solution. And trust me, not all of them goes to product. Most of them are scrapped out. But the thrill that you get in this thinking in advance and visualizing where this whole thing is going to be, it's really exciting and thrilling.

Luke: Just to reflect on that, it seems like it's a perfect time period too. Right? It's enough that you can actually see the results. Whereas if you're in some sort of pure academic research, which is awesome, but it may be decades or something til you actually get the product. Whereas this space you're in, it's closer to what maybe traditional applied research was. Where it's like there's a short enough horizon where you could actually see the results and get things done in a year or two.

Parul Singh: Yeah. That's super exciting. And for me, I would say it's the variety of projects that I get to work on. Just imagine, in a year I move from quantum to edge computing. And that is super interesting. Not many people get to do that. They take up a role, and they continue to do that. It's not like it's not valuable or it's not acknowledged. That is. But I think I really enjoy the pace of work. It's very fast. You have to learn things very quickly, you have to be very adaptable, and then you have to generate code in a very short time. So, that is very exciting for me.

Joe Sepi: I can imagine someone listening to this and thinking how awesome your job is. So, I would imagine people wondering if you are hiring. And maybe you don't know, but I thought I would just pose that.

Parul Singh: They can always come back to our Red Hat Jobs and find if we are hiring, but my team is super interesting. So if anybody's interested, go and look for jobs in the emerging tech group of Red Hat, and do apply.

Joe Sepi: Yeah. Yeah. I looked it up. Redhat. com/ jobs, I think, works.

Parul Singh: Yeah. Yep.

Joe Sepi: And then IBM does a lot of research work as well. So, careers. ibm.com. You can check out there as well. But yeah, I can imagine people listening to this and being like, " I want to do what she's doing." So, that's exciting. That's great.

Luke: I would comment here, that is actually one of my biggest surprises of working in the enterprise, is how interesting it is because of that diversity of things you can get involved with. And I think this podcast is my version of that, where it's, " Oh, wow. This week I can talk to you, Parul, and hopefully we'll have Naeem on later this quarter to hear more about the satellite." So that is one of the things, I think, that a lot of people don't realize. That the enterprise is very exciting because everything's at- scale, it's big, it's real right away, and there's such a diversity of topics and industries.

Joe Sepi: Yeah. For sure. Mary asks a question here that we'll put up onscreen as well. Do you get to choose the topics that you like to work on, and the tech stacks as well?

Parul Singh: So, the topics is usually how it works. The team decides what are the new things that they're going to take up, and then the people on the team decides what they are interested on, and they work on that. But I would say there's a little flexibility, but it has to be aligned with Red Hat's strategy, of course. As long as it makes sense. And as I said in the start, my job is to go and see the emerging technology trends and then find out how it can be aligned to the Red Hat strategy overall. So I wouldn't say there has to be a discord, but as long as it ties down to what Red Hat's strategy is, you do get to choose the topic and the tech stack that you're working on.

Joe Sepi: Yeah.

Luke: It sounds like you already were predisposed to the same ethos of what's going on at Red Hat, having a passion for open source and collaborative working. So, it seems like it's a great fit that way.

Parul Singh: Yeah. Yeah. I love open source a lot, because you get to work with so many diverse and smart people. I believe in the philosophy that you are 25% of what people you're surrounded with, and open source is the best way to surround yourself with really smart people. You do not need to work in the same company. You can be anywhere. I collaborate with people from IBM. When I was working on the Qiskit operator, I get a chance to collaborate with the Qiskit team that is in Spain. I couldn't have conceived that I can do that in a normal job. So, that's super cool.

Joe Sepi: Yeah, very cool.

Luke: We have a bit of a technical question here about edge and how communication... So, how can the data be transferred between services? What type of databus do you implement with events and queues around edge computing?

Joe Sepi: That's specific to VolSync maybe, when we were talking about that, I'm guessing.

Luke: I would interpret this... They didn't say edge, but I would say what comes to my mind, and obviously I want to hear your answer over mine, is MQTT and sometimes WiFi. Maybe if it's really an outpost, maybe it's using Lura. I don't know. So, I think it could be protocols as well as technologies.

Parul Singh: Yeah. I think it depends. For example, when I was doing for VolSync, when I'm working specifically for the edge use case, we are using HTTPS to connect to the intermediate system. But one other scenario uses rsync, and rsync works on the local SSH server. So, it really depends on the use case that you have. But yeah, there's nothing specific. We do use queues sometimes when... One of the operators that I'm working... Not working. I would just say collaborating or consulting the IBM. They use queues. So, it depends on the use cases. I wouldn't say there's one strategy that says, " This fits all." It depends on the use cases, from anything from HTTPS, to gRPC, to SSH. It can be anything.

Luke: All options on the table.

Parul Singh: Yeah.

Joe Sepi: We have another question here I wondered about bringing up. Can an individual earn money from open source? That's a good question. Not an easy one, perhaps, to answer.

Parul Singh: Yeah. So, if they mean an individual who is not working for a company that do open source. I guess that's what they're asking. If yes, then I would say I'm not the right person to comment on that, because I was fortunate enough to find a company that does open source and pays as well. But I recently read this article. So GitHub as a ReadME project going on, where it's shared stories about people who work on open source, and these people are individuals. They do not belong to an organization. So there are scenarios where people do earn money from open source, but it is not a very structured sector. It is little hard for you to be an individual and earn money from open source. Because again, you can have so many people who want to use the project but they are not contributing back. So, most of the time you are an individual who is leading the entire open- source projects. But it's not a hard no. It's not even a hard yes. It's in- between.

Joe Sepi: Yeah. I think in my experience, and I think you were touching on this in the beginning, there are a variety of ways that you can interpret this question. For example, you earn a salary and work in open source, and similar to me as well. And I know a bunch of folks from obviously Red Hat and IBM as well who are paid by our company to work in open source. So, that's one angle. But an individual on their own, that's a little bit harder, but certainly there are people out there that are doing it. I used to work with the gentleman by the name of Henry Zhu, who is the lead person on Babel, a JavaScript transpiler. We worked together at Adobe, and he left Adobe to work full- time on open source, and was, I think, making due. I think he has some articles and does a podcast, so you can probably read more about that there. But I think it's certainly not easy. I think the other angle too is, we had talked about earlier, you can be working in open source and potentially get a job through that work. If you have visible work and you're interacting with folks at a company, there's an opportunity there to get hired as well. So, it's multifaceted question and answer.

Luke: I think these were both great answers. Yeah. It seems like it needs to be part of a bigger strategy, and there's probably room where you could be a consultant around certain things that are important to the industry. But yeah, I love these answers. That it should be part of a bigger strategy of, " This is going to distinguish me in such a space so I can either get a job, or be a consultant, or X, Y, Z." But it's not just, in itself, it's not driving Uber or something. You can't just pick it up and...

Joe Sepi: Yeah. Exactly. Well, looks like we used up another hour. Another great hour of talking through some interesting technology with our wonderful guest. Thank you again for joining us, Parul.

Parul Singh: Thank you for having me. It was fun. So much fun.

Joe Sepi: Yeah. I'm glad you were able to make it. I look forward to working with you more, and perhaps having you on again, and talking some more about some cool and interesting stuff that you're doing in the near future.

Parul Singh: Thank you.

Luke: Thank you for tuning in. If you're catching this as a replay on podcast and you have any questions, please tweet at us. It's been a pleasure bringing you this episode, and see you next time.

Joe Sepi: Yep. Cheers.

Parul Singh: Bye-Bye.