Are you about to head off to college? Interested in DevOps and the Cloud? Is there a good way for someone like you who is starting out in the world of technology to absorb the necessary skills? The Open Source Lab (OSL) at Oregon State University (OSU) is one program that helps students and serves as a career accelerator. OSL is a unicorn because OSU is willing to invest in open source.
Today, we’re talking to Lance Albertson, director of OSL at OSU. OSL does a variety of projects to provide private Clouds that are neutrally hosted on its premises. The lab also gives undergraduate students hands-on experience with DevOps skills, including dealing with configuration management, deploying applications, learning how applications deploy, working with projects, and troubleshooting issues. OSL is for any student who has a general interest or passion for it, and a willingness to learn.
Some of the highlights of the show include:
- Workflow focuses on what students need to learn about Linux and giving access to various repos; then they experience the lab’s configuration management suite
Interview Process: Put out a posting, student submits an application online, each candidate is reviewed, student is given a screening quiz,
If a student passes the screening process, they are brought in for an in-person interview for personality and technical questions
Students tend to initially have the least amount of experience and most difficulty with a repository that has multiple people committing to it and dealing with PRs
Spinning up VMs and understanding how configuration management is connected, how services communicate, and how to set up an application
Round-Robins and System Sprint Meetings: Focus on discussing and documenting processes, issues, suggestions, comments, and other information
Younger students are mentored by Lance and the older students; every generation has to evolve because the environment and industry evolve
OSL made OpenStack work on POWER8, PowerPC, and PowerPC little-endian; gateway into Cloud - having OpenStack instance to offer services
Vast majority of OSL’s revenue comes from donations; no direct support from the university; finding companies to serve as sponsors is beneficial to all
Future of OSL: Providing more Cloud-like services; creating a more internal, private Cloud’ and containerized ways of running or deploying applications
Full Episode Transcript:
Hello and welcome to Screaming In The Cloud with your host, cloud economist Corey Quinn. This weekly show features conversations with people doing interesting work in the world of cloud, thoughtful commentary on the state of the technical world, and ridiculous titles for which Corey refuses to apologize. This is Screaming In The Cloud.
This week’s episode of Screaming In The Cloud is generously sponsored by DigitalOcean. I’m going to argue that every cloud platform out there biases for different things. Some bias for having every feature you could possibly want offered as an added service at varying degrees of maturity. Others bias for, “Hey, we heard there’s some money to be made in the cloud space. Can you give us some of it?”
DigitalOcean biases for neither. To me, they optimize for simplicity. I polled some friends of mine who are avid DigitalOcean supporters about why they’re using it for various things, and they all said more or less the same thing. Other offerings have a bunch of shenanigans, root access, and IP addresses. DigitalOcean makes it all simple, “In 60 seconds, you have root access to a Linux box with an IP,” that’s a direct quote albeit with profanity about other providers taken out.
DigitalOcean also offers fixed-price offerings. You always know what you’re going to wind up paying this month, so you don’t wind up having a minor heart issue when the bill comes in. Their services are also understandable, without spending three months going to cloud school. You don’t have to worry about going very deep to understand what you’re doing. Its click a button or making API call, and you receive a cloud resource. They also include very understandable monitoring and alerting.
Lastly, they’re not exactly what I would call small-time. Over 150,000 businesses are using them today. Go ahead and give them a try. Visit do.co/screaming and they’ll give you a free $100 credit to try that. That’s do.co/screaming. Thanks again to DigitalOcean for their support to Screaming In The Cloud.
Corey: Welcome to Screaming In The Cloud, I’m Corey Quinn. I’m joined today by Lance Albertson who is the director of the Open Source Lab at Oregon State University. Welcome to the show, Lance.
Lance: Hi. Thanks for having me on.
Corey: Thanks for joining me. Let’s start at the beginning. What is the OSL?
Lance: The Open Source Lab is basically the Switzerland of open source hosting. We do a variety of things for a variety of projects from small to large. You can think of us as a co-location facility for some of the larger projects like the Apache Software Foundation, to providing your own private cloud that’s hosted on premises for smaller projects or projects that just decided they want to have a hosting here because we have a neutral way of having things hosted.
Some of those projects include BusyBox, we host BusyBox, Buildroot and all of those various components. That’s what we’re all about. We’re about helping that and then the flip side of it, probably the more important part, is we give undergraduate students at Oregon State hands-on experience doing DevOpsy kinds of skills where there’s actually hands-on on the data center or dealing with configuration management, deploying applications, learning how applications deploy, working with projects, interacting with them, doing troubleshooting with issues, all of that fun stuff that you really can’t get when you’re in school.
Corey: What’s fascinating to me about it, and I’ve talked about it on this show before with various guests, there’s no good way for someone starting out in the world of tech today to absorb all of these things. You talk to people who are—in my cohort, where 15 years ago we’re looking at a radically different world—the path that we walked tends to have closed behind us. There’s now not a great series of answers to how do I become you? How do I get to a job where I get to work with these things starting from scratch. The OSL is one of those programs that I have been aware of for a while that tends to answer that question very efficiently.
From the student side, when someone enrolls at Oregon State University, how does becoming involved with the OSL manifest itself? What does that look like?
Lance: It depends on the student. We have had some students specifically come to OSU because of the OSL but we’ve also had some other students that never heard of us until they were maybe a lab meeting here on campus or they had a friend that worked at the OSL or know somebody at the OSL, they hear what we do, and like, “Hey, I’m interested and how do I get involved with that?” That shows you like word of mouth of coming to understanding that.
Corey: Are these students generally enrolled in a particular program? Is it tied to a particular discipline at the university? Or is it more general, in that, anyone who has an interest?
Lance: I think anybody that has a general interest, has the passion for it, and are to willing to learn it. Historically, most of our students have been in the CS side of things, but we have had students that had even anywhere from a range of the physics students, we even had mass students, but we’ve also had one student who came back from his previous IT career, coming back to OSU, enter their fermentation science program, open to us as well. We have a wide array of students that are involved here, but a lot of them has some kind of a CS background.
Corey: Someone winds up enrolling in a program, they were getting hired by the program since this does pay them, what does day one look like?
Lance: Day one, just like anything, we create all the accounts, we make sure you can log in, we give you keys and all that, but then the next step is we have an onboarding process, kind of a new hire checklist. We’ve worked out perfecting it. It’s nowhere perfect still, but we we keep improving upon it each time we have a student go through. It goes through having to learn about the infrastructure, about what we do.
A lot of these students just dabbled in Linux so we have a workflow of here’s basic things you need to learn about Linux, kind of a crash course on that. Then they also work with other students that kind of do that. From there we dabble with them, giving a little bit of access to the various repos, having them go through the workflow of getting approved on things, reviewed. Then we dive in to our configuration management suite, which we use Chef primarily. We dive right into Chef and Ruby, and all of that. They obviously won’t do that on the first day, but that’s where we lead them off at the very beginning.
Corey: What is the interview process look like for hiring someone? Historically, when I’ve been in operation management roles, it’s easy. Just find people who’ve done all this stuff before. When you’re hiring undergraduates who never held a job, in many cases that didn’t involve deep fat frying or mopping things, you can’t necessarily find that basis of experience. What do you look for when you did this thing?
Lance: That’s been the work that I think we've kind figured it out over the years. What we do is we put out the posting and then they submit their application online. They remember all the process as we go through. We review each candidate and we see where they’re at. We don’t look at their technical credentials as much as if they at least have some experience that surely what we want.
From there we will send them the screening quiz, open book, open test, that’s just kind of some basic stuff. We have them send that to us in a plain text file that we put in a git repository and we all internally review it and say, “Hey yeah, they seem like to understand what this concept is.” It goes from very basic stuff to maybe we have him do a little bit of a Chef thing. This weeds out those people that heard about it, maybe they don’t want to go through the whole process of learning about it. We need to know that.
They go through that screening process and once we determine that, then we bring them in for an in-person interview. The interview, we used to have it for an hour, but that was just way too long for this type of a thing, so we tried to pare it down to just 30 minutes.
I’d say about a third of that interview or maybe even half of it is just getting personality kind of questions, understanding what they’re interested in, where they’re going, what they want to do, do they fit in what organization you know, and all of that, and then we start diving into technical questions.
I always tell them when we get to that part like, “We don’t expect you to know all of this stuff. If you do, you’re a rock star.” But just think about it, we mostly look at their problem-solving skills and we help them along if we get them to answer something, or not quite get there like, “How about this instead, or that?”
We try to do a variety of systems and a little bit of programming questions. We used to do some kind of a programming quiz, but that really didn’t help what we needed. We couldn’t figure that out yet. After that, it’s just the termination of based on all those things what do we do? The nice thing about all of that is I have my current students actually in there being part of the interview as well. They get to see what it looks like on the other side of the table, so they learn that as well. That’s what it’s all about.
Corey: Do you find that there are certain concepts of modern operations––DevOps, CloudOps, SRE––whatever we’re calling it in the 20 minutes between the time we record this and the time this winds up going out, that students have more trouble with than others?
Lance: What’s that question again?
Corey: As you’re bringing people onboard and exposing them to a whole raft of new concepts that are central to the ops, DevOps, SRE, whatever we’re going to call it in the 20 minutes between now and the time this publishes—there will be four new terms for it—of the entire arena of things that go into that space, are there any particular areas that you see that is more difficult for people to wrap their heads around than others?
Lance: Let me think about that because I think there’s a variety of things and it depends on the student as well. The first initial thing that a lot of these students don’t have a lot of experience in is dealing with a repository that has multiple people committing to it and dealing with PRs.
Most of the time, they've have some kind of a school project, it’s usually just been them probably pushing to master and so forth. Now they actually have to go through a review process. Then some automated test comes back and it says something failed and they have to fix that and figure that out.
Usually, the first thing I’ve noticed the students have to figure out and then beyond that it’s just kind of understanding how all of this works together like, “Oh, this is a VM running running on a machine that’s doing this that has a storage connected in the back end,” and just understanding how it all fits together.
Beyond that, it’s understanding how the configuration management ties into everything, how all the services talk to each other, when you set up Apache or whatever, how do you actually set up application to do the thing. It depends on the various students and what happens. We’re not doing massive things like we’re doing in the public cloud. We’re just spinning up VMs basically. It’s a basic thing, so understanding how that all works.
Corey: Which is very fair. To that end, something that’s fascinated me about the OSL for a long time is you first talked to me about this when you were hosting some stuff for the Freenode IRC network where I used to be staff. I think that’s how we met about 10 years back.
In that time, it was a very different world back then. Even VMs weren’t so much of a thing back in those days.
Now, not only we gone past VMs, you mentioned earlier when we spoke previously that you’re doing things with OpenStack, you’re exploring things in the public cloud. What’s fascinating is that this pattern tends to manifest itself all over the place in large enterprises as they’re going through their own digital transformations.
What’s neat about the OSL, comparatively, is because you’re staffed with undergraduate students, on day one of bringing someone aboard, there’s an expectation that in several years regardless, that person is not going to be there anymore, full stop. You’ve been able to execute a transformational series of changes that have all been bounded on day one by the knowledge that this information can’t live locked away in one person’s head.
How have you found that that manifest itself as far as how you approach documentations or how you approach articulating tribal knowledge and recording that in a way that someone will be around to finish a project when the people who’ve started it have long since departed?
Lance: That’s a really interesting question. It’s something we’ve had to tackle over the years. I think we’re still trying to come up with the best way of doing that. I think the biggest change that’s really helped us is in years passed, we had students that have a specific area they worked on. They would work on that, it was great, everything was awesome and then when they left it’s like, “Oh crap. They didn’t maybe document everything. How do we do that?”
More recently, we’ve switched to where we do a round-robin of where everybody gets to understand how all the things work with all the stuff the best they can. They still can specialize here and there but they get to do that. We have what we call a system sprint meeting. Having a sprint meeting with ops people is kind of odd.
Basically, we just have this trouble board and we have various top-level things we want to work on and we have columns where we’re currently testing and researching it, moving on, moving forward, we go through all of that, and so we’ve kind of even done it with documentation of it. “Okay, we need to document this now.” I assign this person to work on it and do that. We try to make sure everybody documents to the best they can everything that’s going on.
We internally use Sphinx stack for all of our documentation. We have everything in a git repository, it’s all reStructuredText, so you can use your editor and edit things, then we just put that in an overquest, and we can get reviewed, people can see. We make any documentation fairly easy for a lot of these things.
A lot of the students that run into issues were like, “Man, I wish it was documented better.” Half the time it’s just me not actually documenting it that well myself, unfortunately. I’ll teach them around on what’s going on and they go through it, may learn it, they actually document it, and go through that.
It’s getting everybody to know everything and then there’s this mentoring process where the younger students get mentored not only by me, but by the older students, and the older students help trickle that knowledge down on each generation as they go through it, kind of cycles and does the same thing, and goes on through.
Corey: A counterpoint, too is that, if you take a look at where this started and where the world has evolved, every generation of student has to have evolved because the environment has evolved and the entire industry has evolved.
Back when I started working in operations myself, one of the rainmaking skills that changed everything was knowing the ethernet B standard where we made our own patch cables. It turns out that that doesn’t pay the bills quite like it used to in a cloud-first world. How has the program changed as the industry has changed?
Lance: We’re still evolving. What we’ve been trying to do is deploy more cloud-like services for a lot of these projects and trying to get our students more involved with that, whether that means embracing using Docker in places, using OpenStack in places, just having a more open-minded approach of how we deploy applications, how we manage things. On the flip side, the projects have to be that open-minded as well.
We try to embrace as much cloud-like technologies as we can. The first stuff we did was dive into, looking into any services more API-driven. Historically, and we still use it, we use most of our virtual machine technology, we use something called [...], which is a project that came out of Google. This project came up way before OpenStack even was, but it’s not very API cloud-friendly. It’s good for just, “I want to VM that’s up, it’s always up, and I can manage the way I want it.”
We’ve had a lot of things come up where we needed to have a service that’s self-serving, that gives projects what they need when they want it. We finally started embracing OpenStack and getting that deployed. Beyond that, once we got that deployed, we realized we need to have a better way of handling storage in the cloud than how we manage things in OpenStack. So that meant we need to learn about Seth and getting that deployed, and then all the networking behind all of that as well with Neutron. When we switched to OpenStack with Neutron and all of that, that was a big thing.
Probably one of the big success stories we’ve had more recently is a project we've had with IBM. For over 10 years, we've hosted a power server here for projects, but it’s always been we give you shell access or we spin up, God forbid, an LPAR using their arcane technology to make that work.
Corey: Now that’s a name I’ve not heard in a long time.
Lance: Yes, LPAR is an HMC. I’m still dealing with that somewhat, but we’ve been setting up LPARs for projects so they can build stuff on POWER. They came to us and like, “Hey, we want to expand this and make this work better on POWER8 when it came out.” They completely changed my systems worked on POWER8. It actually boots up like an Intel machine almost. It actually boots up normally somewhat. I can use IPMI, for example to connect to it.
Anyway, the only way we can do that is making it cloud-like. So I worked with IBM and their engineers. We were leading edge making OpenStack work on POWER8, PowerPC, and PowerPC little-endian and getting that going. Now that we’ve done that, we have, I think, almost 100 projects that we have that are hosted on that platform. That really gave us that feel of OpenStack is working. We feel that’s pretty good, so now we just recently finish up spinning up our x86 cluster for just general projects. We've just started having some projects on there and getting on there.
My hope is that’s our gateway into cloud, so to speak, is that we have the OpenStack instance and we can offer some of the services. I feel we’re more robust that we can provide some of that. I’m also working at having some containerized service that goes with that as well, just trying to see what we can make it do.
A lot of it is just getting anything we can have, be API-driven as much as we can because like any IT infrastructure that’s been around for more than 10 years, we have a lot of cruft of things that need to be fixed. We’re still trying to work through a lot of that.
Corey: A common and somewhat myopic view of folks who are involved in cloud-native environments, and I’m very frequently guilty of this myself, is to discount anything older than about four days as legacy and not worth paying attention to. If it works, it’s not interesting. That is a mistake in that what you’re describing is, in many ways, an exact one-to-one comparison with what many large companies who are not traditional technology companies, are working on and how they’re proceeding with things.
It’s good to remember that this is a very common story for many of our listeners as well as for many of us who leave one company and then find out that, when we’re no longer working for a multinational tech company, that spending extortionate piles of money on infrastructure first, is not somehow on the strategic road map. Being able to look at things that have been working for a long time and be able to address those is incredibly valuable.
To that extent, some of the most impressive technologist that I’ve met have been OSL alumni. Do you have any examples of people that you can mention who’ve gone through the program and what they’ve gone on to do?
Lance: I think by far the most famous example is probably Alex Polvi and Brandon Phillips. They were leaving here as I was starting here. They were students here at the Open Source Lab when we also actually started. Alex, he first created, I can’t remember the name of that company now, but it was bought up by Rackspace. Then he created a second startup company CoreOS, which a lot of people know about, that recently now got bought by Red Hat. Those two alumni came through this program. They have that feeling of what really needs to happen.
Beyond that, we have folks like Emily Dunham that has gone through this program. I can’t name all of the people we have but were scattered throughout the IT industry. We had people, whether at Red Hat, Ansible, Microsoft, actually. One of our students has been at Microsoft for quite a while and she worked on a lot of the Hyper-V stuff and probably wore her Mozilla t-shirt there in the early days.
We’re out there and everywhere. It’s a great community hearing where people have ended up, what they’ve done and the experience and the opportunities they got. I almost feel envious because I wished that was me when I was at that point when I started out, but it’s giving me joy seeing where they end up.
Corey: Why is it that the OSL is one of the only programs like this in the world? Back when I was coming up, I would have given a limb to work on something like this. It would have given me such a leg up on this entire industry. Why is this a unicorn?
Lance: I think there’s a lot of things unique to us and just was the perfect time, the right time. When the OSL started, it was the post-dotcom boom in the early 2000s. At that point, a lot of open source projects didn’t really have a lot of places to have their hosting, so we had, at that time, there was some dark fiber going up and down the valley that we could tap into and actually utilize all the bandwidth. Back then, bandwidth was really hard to find for a lot of these projects, so we helped with that.
At the same time on the flip side of it, the organization here at OSU, they’re willing to invest in open source. They had their own mini transformation in that time going from proprietary to using more open source. The CIO at the time, he was really supportive of us and top of the university president. That really helped us in getting up. A lot of it was just trying to find that money and right up that point, we started actually getting word-of-mouth and then we got some initial funding. Actually, do you remember Real Networks?
Corey: Hang on. I’m trying to remember that but I’m still buffering.
Lance: Yeah, exactly. Back then, some of our students actually worked on adding some applications to the OLPC program that had come out, and so we worked with Real Networks to actually port some of the streaming stuff over at that point. They liked us so much that they gave us, I think, it was like a half million dollar donation that was able to get us going at that point.
That got the answers to some other companies like Google and so forth to kind of see what we’re doing and how we’re providing all this infrastructure and everything to the lab and everything.
If you do it at another university, it’s going to be really difficult because of university politics, funding. All of the funding we get is through funding through either donations. We do have some hosting contracts that’s mostly for the larger projects that we host that are able to pay for it, or they want to pay for it in some way, but that covers that cost, but the vast majority of our revenue comes from donation. We don’t get any direct support from the university.
If you try to replicate this anywhere else, that’s usually the showstopper. You need a data center and you need to have money to pay for all this stuff. Then you need to have staff that has all this knowledge and know-how of how to do it.
Then having the right people above and below you to work on all of that. It’s kind of a perfect storm to make that happen here. I don’t know why it really hasn’t happened to other universities, but that’s how it started here.
Corey: Absolutely and finding companies to sponsor stuff like this is incredibly important. If nothing else, just from a perspective of, “Well, we all got to where we are in this industry because we stand on the shoulders of giants and people did favors for us. How do we help the next generation?”
But there are also tangible benefits to companies for being able to get involved in stuff like this. Just from a perspective of publicity, of getting a pulse into what it’s like for people who are starting out new. Are there any current sponsors for example who’ve done interesting things in the context of leveraging their relationship with you folks?
Lance: Some of our sponsors are pretty much hands-off. They’re just like, “Hey, just keep doing what you’re doing. You’re doing great. We’re keep sponsoring you. Just keep offering free hosting to projects and doing that, and that’s great. That’s all we need to know.”
Other sponsors like IBM actually have an interest in having people get on their platform, switching it to running on their architecture. That big behemoth that’s really hard for them to say, “Hey, come over here on our thing and have access to it.” They see the value of coming to us and they have the name, they have the relationships with the open source projects, they have the trust, let’s utilize that and then we can maybe help get things going. That’s probably the most hands-on group that we have as well.
We’ve had some experience with Intel in the past, but that’s an on-again-off-again thing. At one point we were hosting MeeGo, for example, which turned into Tizen, and all of that. That was a mobile platform that I think Samsung or somebody eventually started using. But we did some of that.
With Facebook, we had some connections back in the Prineville data center here in Oregon where they have their first big data center they created. I actually knew somebody who used to work at the state of Oregon that got a job there and they were able to actually send us some hardware from over there. We actually have three racks here of their pre-production hardware that they run at the data center. That’s the Open Compute architecture so we were able to get hands-on exposure to that.
That’s one unique area of like, “Hey, we can actually maybe use this and give the exposure back to them.” So it really depends on the company. A lot of them are hands off or they've been hands-on.
Corey: One of the more memorable interactions we had was a few years back when I was running an ops team. I asked you whether any of your students who were graduating soon would be interested in having a chat about potentially coming to work on my team. It’s a testament both to how well-mannered you are and the strength of our friendship that you didn’t laugh in my face when you said, “That’s adorable, get in line.” What astounds me is what a career accelerator it is for students who go through the program. I don’t think I’ve ever heard of someone who was in the program looking for work for longer than about 20 minutes or so.
Lance: That’s certainly is the case. A lot of these students, they set up and they have an internship at some point in their sophomore, senior year and that’s usually when they get that hook in because they’ll come back. If I know they’re going that summer somewhere, I’m going to say there’s a 50/50 chance that they’re coming back and they’re going to work the rest of the summer for me or they’re going to come back and they’re going to maybe work part time remotely for this other company, they finish out school, and they may already their jobs set up before they graduate.
It doesn’t always happen that way, if a student––I guess, if they were really lazy and they didn’t actually do any of that, then maybe they would have a hard time, but that’s usually how things go. They usually have a lot of options.
I remember I had one student that kept going to a variety of companies, having an internship, and he’d already been here for five years. I said, “Dude, graduate and get a real job,” because he’s having so much fun doing all the various things. It’s incredible. They have a lot of opportunity where they can get started.
Corey: As you look back in the past decade and all of the changes that have happened to the OSL, what do you see when you look forward? What does the future look like for you? For as you see the world continuing to evolve in a more cloudy direction, as you see the skill set of the people you want to hire continues to evolve, what changes are on the horizon? What do you see coming?
Lance: This is something that I have pondered a lot for the past year, year-and-a-half, and it’s a really difficult question to kind of answer and I think the short answer is trying to provide more cloud-like services for projects to use.
A lot of what I see projects coming as what they need is a little bit more robust CI environment that can run long jobs or larger jobs that maybe they can’t fit on Travis CI or whatever CI thing that you have running.
We have enough hardware for the most part. It’s just having the right amount of services set up for those types of things and getting that going, and so kind of getting that going and creating more of an internal private cloud for that type of stuff. I’ve been diving into more containerized ways of running or deploying applications.
I see us probably eventually diving into something like Kubernetes for maybe managing some of the applications that we host for projects. That would be a great learning resource for a lot of the students here, so they can get a head start.
The thing I’m trying to do is make sure that my students have enough modern level of that information and knowledge, so that when they graduate they are employable. Usually, it’s not a problem, but I don’t want them working on things that, “Oh yeah, we did that 10 years ago, so we’re still going to do that. Now that’s the way we do it.” I keep trying to evolve and being a sysadmin myself, I find myself cringing if my old curmudgeon sysadmin ways, “My God, can’t do this,” but I have to be open and just like, “Oh yes. Let’s do this.”
Software defined storage is a thing and it does work. It’s not completely slow, as long as you do it right. That kind of stuff. Just finding flexibility in how we deploy services, and trying to move more to ephemeral type ways of hosting applications for open source projects.
I’ve heard a lot of projects having interest in running a local GitLab instance, for example. Maybe they don’t want GitHub or something. We could be a resource for that. Whether it is having the bigger SCI experiment or having access to the “exotic architectures” we have current access to POWER.
I reaching out to ARM to maybe do something similar with them. I’ve also worked out with trying to see if I can even get RISC-V equipment as well. We have the space, we have the hands-on people, we can deal with these types of things, and we can give that access.
Corey: It’s really neat to see the ongoing evolution of education in this space. There is most assuredly more than one way to do it, but it’s very obvious that you found a way that works.
Thank you for taking the time to speak with me today. I deeply appreciate your time, Lance Albertson, Director of the Open Source Lab at Oregon State University. I’m Corey Quinn and this is Screaming In The Cloud.