DevOps Institute

[E19] Embracing the Concept of Liquid Software

Podcasts

Baruch Sadogursky, Head of Developer Advocacy at JFrog and DevOps Institute Ambassador talks about how organizations can benefit from liquid software in times of global crisis.

The lightly edited transcript can be found below.

Intro:

You’re listening to the Humans of DevOps Podcast, a podcast focused on advancing the humans of DevOps through skills, knowledge, ideas, and learning, or the S-K-I-L framework. Here’s your host DevOps Institute CEO Jayne Groll.

Jayne Groll:

This is Jayne Groll, CEO of the DevOps Institute. Welcome to another episode of the Humans of DevOps Podcast. I’m really delighted to be joined by Baruch Sadogursky, Director of DevOps Advocacy. I hope I got that right Baruch, with JFrog. Good friend, a good advocate, and now at DevOps Institute ambassador. Welcome Baruch.

Baruch Sadogursky:

Thank you. Thank you, Jayne. Yes, super, super excited to be the DevOps Institute ambassador. That’s a great honor. Thank you for accepting me to the program and super excited to be on this podcast. Thank you for having me.

Jayne Groll:

I’m excited to have you too. Baruch, as you know, we’re in the middle of a terrible global crisis and this pandemic … This is the Humans of DevOps Podcast. This is a human crisis. And I think it’s taught us some lessons that, despite the fact that nobody predicted this, nobody wanted it, but I think for enterprises and for individuals, we’re learning some pretty hard lessons from this pandemic.

Jayne Groll:

I want to talk to you a little bit about the people process and technology lessons that we’re learning when most organizations around the world have had to pivot very quickly.

Jayne Groll:

But before we begin, I know that you and your colleagues at JFrog wrote a really great book about liquid software. Can you tell us a little bit about what liquid software is?

Baruch Sadogursky:

Yeah, for sure. Liquid software, you can learn more@liquidsoftware.com obviously. It’s a concept that talks about a next step of continuous things. You have continuous integration, you have continuous delivery, you have continuous deployment. And the next step for us, at least how we see it in general, is the continuous updates.

Baruch Sadogursky:

Continuous updates are the realization that what we do in all those continuous things is actually more often than not updating the existing software instead of installing a new one, which basically means two main things.

Baruch Sadogursky:

The first is, your updates are small and frequent. As opposite to when you install a new application, it will be a big and rare event. And your updates are updating something existing instead of setting up something new.

Baruch Sadogursky:

When you start thinking in continuous update metaphor, you will actually realize that the more frequently and the smaller you have updates, the better it is for both the producer, the vendor and the consumer, the user, the customer.

Baruch Sadogursky:

If we take this idea of small and frequent updates to an extreme, it actually becomes a flow of very frequent and very small updates, up to a point when software looks like liquid that streams from the vendor to the user. This is where the liquid software metaphor comes from.

Jayne Groll:

I love that, because I was an IT director in a galaxy long time ago, far, far away, and the concept of patch Tuesday. Where every Tuesday, you took all the patches you got from all your vendors. I’m sure you lived that life at some point as well and the risk factor. the risk factor was tremendous.

Jayne Groll:

I mean, on the op side, there was always the concern that, which one was going to cause an outage. And if an outage occurred, how did you know which one it actually was? Right? And in today’s very fluid environment, I love the analogy of it being liquid. Because again, the idea of being continuous, liquid flows until it runs into some type of a barrier. And it sounds like removing those barriers, going small and more frequent, which is really the kind of the core of DevOps, is something that’s particularly important.

Baruch Sadogursky:

Absolutely. Yes.

Jayne Groll:

I started by saying that this terrible situation has also taught many, many organizations lessons, some hard lessons. Some really hard lessons, in terms of dependency on technology. Organizations found out, some very well, some very poorly … By the way, if you’re a COBOL programmer, you have some enormous opportunities ahead of you … but it also showed that it’s no longer an option, I think, to be able to look at decoupling, to look at moving to the cloud, to look at how you build your applications and what artifacts and assets that you use to do that.

Jayne Groll:

I think, more important than anything else, it showed that humans are probably the critical success factor that will make or break your ability to adapt to changing business environments. And in this case, many organizations had to pivot very, very quickly, often in a matter of a few days from an on-prem, in-office environment to a off-prem, work-at-home environment.

Jayne Groll:

So I wanted to just chat with you a little bit about that. I mean, obviously, the work that you do, you’ve been an amazing advocate for the community. JFrog, I think plays a really important place, in not only the DevOps community, but in the enterprise community.

Jayne Groll:

Liquid software, how can organizations embrace the concept of liquid software and look at what they were doing before, how they’re adapting today and potentially what do they need to do differently in the future? So I know that’s a really big question, but maybe you can break it down for us.

Baruch Sadogursky:

Yeah. Yeah. Yeah. I think I can tie into couple of things that you said and bring it back to DevOps and liquid software. And I wanted to start with the COBOL example because it’s so amazing and so profound.

Baruch Sadogursky:

So the most interesting story about this COBOL thing is that it proves how much liquid software can save you from suffering an unnecessary expense in completely unpredictable scenarios.

Baruch Sadogursky:

So what we are talking about, why suddenly the COBOL developers are in such a high demand, it turns out that a lot of systems that manage unemployment payments were written in COBOL and never bothered to be updated for one of the reasons no one thought it’s necessary.

Baruch Sadogursky:

For many, many years, the unemployment was low. The systems were adequate for the load that was at the time. But then, because of the pandemic, obviously the unemployment skyrocketed and those systems didn’t manage to take the load of all the people that applied to unemployment benefits.

Baruch Sadogursky:

This is where, now, government is looking for COBOL developments in a very panic mode and ready to pay any money actually for someone who can actually fix those systems. And the problem here is, amazingly as the problem in a lot of other aspects, is a lack of updates. Those systems were never updated because people thought, “Why bother?” And here’s an example for you. If you don’t update your systems for 20 years, when something happened, you will be in a very bad situation.

Baruch Sadogursky:

If this system were updated continuously … I’m not even saying liquid software, but that would be ideal, but even more frequently than once in 20 years, they’ll probably use more adequate technologies today and there will probably be a more robust and more elastic and more able to stand this peak. And obviously reply and be able to handle the situation without now trying to hire developers, which are in very low supply, for any money and suffering, obviously huge losses in terms of both uptime and service and in end of the day money that they need to pay to those rare unicorns that can fix the system that wasn’t updated in 20 years.

Baruch Sadogursky:

So basically here you go. Something that you say, “Well, I don’t need the updates. No one wants to hack to it, so there is no security needs. We don’t need new features because it didn’t change. So why would I bother?,” and suddenly, you see how even here updates are critical in order to prevent this kind of event.

Jayne Groll:

And at the end of the day, while they think that they’re saving money by keeping a system stable, no cost, no risk, at the end of the day, high cost, high risk.

Baruch Sadogursky:

Exactly. Exactly.

Jayne Groll:

Right? It doesn’t take a pandemic for that high cost, high risk to suddenly become necessary. Right?

Baruch Sadogursky:

Absolutely. They might have any other kind of failure that wouldn’t be able to fix because no one touched this system in 20 years. Absolutely. For sure.

Jayne Groll:

Yeah. It’s scary that way. So that’s where they are today, and it’s interesting that it isn’t just one state. It’s multiple states that are struggling with this, “Man. I wish I had learned COBOL.”

Baruch Sadogursky:

Well, I wish they would do liquid software.

Jayne Groll:

Yeah. Exactly. We know what the past looks. We know what the present looks like, at least for those operations. And I’m sure there are similar stories outside of the COBOL dilemma, where organizations didn’t take advantage of some of the innovations. And we’re not even talking about being a unicorn. We’re talking about some of the opportunities, the advocacy for breaking down into microservices, containers, things like that. That’s not really new. I mean, what, five years that haven’t been taken advantage of? So what do you think the future looks like?

Baruch Sadogursky:

Yeah. Yeah. Before we go into the future, another example of today’s situation that also kind of magically ties to … or not surprisingly ties to liquid software, is the disruption that the pandemic applied to entire industry. Suddenly, the things that we thought are less important become the most important things.

Baruch Sadogursky:

Obviously all the remote work collaboration software, like Zoom and again, their ability or inability to react to the market, to sudden market awareness of them. And obviously all the criticism that followed them when people started to pay attention and saw the problems. And here again, how fast a company can respond is basically in the heart of how frequently and how well they can update their software.

Baruch Sadogursky:

I feel for them because you can take any company now, the best unicorn in the world, and if suddenly the entire world, all they’re doing is they’re going through your dirty laundry, they will find stuff because no one is perfect.

Baruch Sadogursky:

This is why I feel for them. Because it’s not that this Zoom is horrible in terms of securities privacy. No, they are average. It’s just suddenly, they are the talk of the day. Everybody are looking at them and finding stuff that they overlooked in Zoom and in other tools that they just didn’t pay attention to.

Baruch Sadogursky:

So the question is not how zoom could prevent it. Probably they couldn’t because it’s written by humans, mistakes are made, software is hard and everybody have problems. The question is, okay, now you are in the spotlight. How fast can you ship those updates that will fix the problems? And you will prevail as a hero instead of being the joke and getting laughed out of the industry.

Baruch Sadogursky:

Again, we find ourselves with liquid software in our hands, that obviously should fix this problem as well, because the faster Zoom can ship their updates, the better shape they will be. And the response time to those comments, to those accusations is the critical piece of Zoom’s survival.

Jayne Groll:

Well, and you know what’s interesting about that is, you’re speaking about it and unfortunately we even coined a new part of our vocabulary, Zoom bombing. And it’s like, “Really?” Within a week, we come up with new terms and that’s a little bit scary.

Jayne Groll:

But what also I think is not recognized, that when we look at continuous updates, updates historically have a lot of toil associated with them. It’s a laborious effort, where we have to build, we have to compile, we have to test, we have to …

Jayne Groll:

In the old days, we’d do releases, what, once a month, once a quarter, once a year?

Baruch Sadogursky:

Yep. Yep. Yep.

Jayne Groll:

Right? Now, the small, infrequent … We have to look at the underlying message, small. Right? First of all, small.

Baruch Sadogursky:

Exactly.

Jayne Groll:

Right?

Baruch Sadogursky:

Yes.

Jayne Groll:

So frequent good, but small. Right?

Baruch Sadogursky:

Yes. Yes.

Jayne Groll:

Because small is manageable. So, little drops of updates, as opposed to the big bang, higher cost, higher risk. And so hopefully coming away from this, the ability to adapt to that, if you’re small, you can be frequent, but if you’re not, you can’t be.

Jayne Groll:

I think that, also, the human side of it. That when we look at what the human toll is … Now in a pandemic, obviously we’ve learned a lot about the need for humans, which sounds ridiculous, but it’s true that in an automated world, we need humans. So, looking at kind of the liquid software approach, what element of that is human and how does that support human transformation, making humans more productive? Or in this case, allowing humans to focus on things that truthfully, today at least, only humans can do, which is really the critical thinking element. So how does liquid software support the human aspect of software delivery?

Baruch Sadogursky:

Yeah. The entire relationship between humans and machines and robots, in the day of accelerated software delivery is fascinating. Because what we actually say is that, in the end of the day, when we do those continuous updates, when we build those pipelines, we want to automate people out of the process because people also make mistakes and obviously cannot keep up with the velocity that we expect from continuous updates and liquid software. But obviously, people are not replaceable at the moment.

Baruch Sadogursky:

What we advocate for is shifting the responsibility of humans from validating the software to validating the pipeline, because validating the software means we have a hard stop. We have a manual validation step in which people will look at the software from different angles. It might be QA, it might be security, it might be performance or whatever, and then kind of approve it to go to the next step in our pipeline. And instead, we need to move people to build and validate the pipelines, so those pipelines can then apply the attestations that human intended to the software, without slowing down the flow of continuous objects and the liquid software.

Jayne Groll:

You know what’s interesting about that, is that humans have the end to end. We can see. We can see the end-to-end process, hopefully. I think one of the reasons that agile software development struggled to really kind of gain the foothold that it was intending to gain was that yeah, the software development teams were starting to iterate. They were starting to go smaller and more frequently. They were starting to iterate, but the downstream activities were still pretty slow because they were very manual.

Jayne Groll:

I think that as we start to look at, particularly for developers, having that kind of core vision of the pipeline, instead of just this kind of siloed blinder approach, is something that we’ve seen a lot of growth in, I think. But I still think we have a long way to go.

Jayne Groll:

And again, I go back to where we are today. I think that there have been some really hard lessons learned about even simple things, even simple things.

Jayne Groll:

I mentioned before we started recording that my husband is working from home. He works for a fairly traditional organization and they’ve got a long legacy and they’re very successful and all of that, but they’d never used WebEx before.

Jayne Groll:

So now, suddenly, you’ve got a 200 person leadership team that has to learn how to communicate without walking to somebody’s office or going into a conference room or getting on the telephone or anything like that.

Jayne Groll:

Now, very simple example, but suddenly take that organization and multiply it by, I don’t even know what the factor is, and suddenly, these software packages are not elastic enough to be able to handle it. They’re busy, struggling to put in some kind of an update. Their team is now working from home, so they’re remote. So, what was originally a very disruptive event, which it still is, layers of destruction kept being added and added and added.

Jayne Groll:

Now the good news is, I’ve spoken to several CIOs recently who have said, “It was horrible, but we’re okay now.” So, they’ve learned to adapt, they’ve learned to work around. But they’ve also said, “Yeah, we’re okay right now, but we know what we need to do. We know maybe not what we need to do, but this is no longer optional for us.” It’s a mandate. It’s a global mandate. It’s not even limited to one industry. It’s a global mandate. What’s your experience with that?

Baruch Sadogursky:

Yeah, no. I mean, that’s exactly that. People have to adjust. And I think this is what we have one of the last advantages to the machines that do a lot of stuff better than us, more accurate than us, faster than us, but obviously the adjustment is where humans excel. This is why we survived all that long … well, until 2020, at least. We’ll see what happens next. But absolutely, this is where humans should apply their next evolution after what they figured out, after the next big disruption, like now, to those pipelines and to those machines. Absolutely.

Jayne Groll:

I do think there’s another message that comes through. You and I talked about this before. The one message I think that’s being carried just across the globe is, take this time to grow a skill. Because coming out of this, and you know this, DevOps Institute does an annual Upskilling report. So, looking at which skills are considered must have, nice to have and not as important, has become clearer year over year.

Jayne Groll:

We just finished our second year, but this is a good opportunity, so that we prepare the necessary skills that, when we’re out of the pandemic … And I don’t think we’re reverting back. Everyone is talking about, what does the new normal look like? I don’t know the answer.

Baruch Sadogursky:

Exactly. Yes. Yes. Yes.

Jayne Groll:

But we know the new normal is going to require new skills. So, if you’re listening, I think that there’s a clear message to individuals and hopefully their organizations support that, that this is as good a time as any to start identifying which skills. Which skills do you think that you need?

Jayne Groll:

So in the arena of liquid software, if somebody were going to focus … Say I’m a developer. So you’re in charge of developer advocacy. What skills would you suggest that they start to grow?

Baruch Sadogursky:

Yeah. So first of all, absolutely. You’re very, very right. We will have a new normal. All the digital transformation that we now experience, which is forced on us because of the consequences, is not going away.

Baruch Sadogursky:

Some topics that dear [inaudible 00:22:32] of us, for example, the tech conferences, IT conferences, they will be disrupted forever. Even after we have the herd immunity and we think we can safely go back to our beloved gatherings, the online conferences are not going away. We actually gained a new format because of what’s happening now, that will stick with us after everything is done.

Baruch Sadogursky:

F you now learn how to run this kind of conference, how to speak at this conference, how to attend the conference without being distracted by your fridge or your TV, and actually be able to participate in it, those skills you will carry out when everything will get back to our new whatever it will be.

Baruch Sadogursky:

Today especially, it’s an amazing opportunity to learn new skills, both because some of us are forced to … because of the new reality, we cannot just do the things the old way, and the others now have the free time on their hands to actually put it in a good use. So, that will be an amazing to actually up your skills.

Baruch Sadogursky:

I think that the skills that DevOps forces us to learn are not only less, but even more relevant today and especially when it comes to collaboration. That’s because the collaboration now, it’s harder because it’s remote. And that’s in the heart of DevOps. Right? DevOps is about collaboration. So, how can we collaborate more effectively when actually the reality kind of makes this collaboration harder? I think this is one of the most important questions of today.

Jayne Groll:

I agree. And actually in the Upskilling report, one of the categories of skills is what used to be called soft skills. We’ve now reframed as human skills, and collaboration and communication, absolutely number one. Absolutely agreed across all layers of the organization, that being a better collaborator …

Jayne Groll:

And by the way, that’s not necessarily natural. It’s not like you’re either born a good collaborator or you’re not right. There are some very proven skills and exercises and ways to just practice being a better collaborator, that’s more intentional.

Jayne Groll:

And then communication skills. There’s the saying, if you think you communicated, double your efforts. Because again, what you say and what people hear are different, but they are different from each other. Right?

Jayne Groll:

Communication is, in some ways, more passive. I tell you, you tell me. Collaboration means I respect your expertise and I ask for your opinion.

Jayne Groll:

Empathy really rose up as a key skill, putting yourself in the position of somebody else, and I think that’s important as well. And I appreciate that Baruch, because I think too often, if I were to ask that question, they would say, “Well, learn Kubernetes.” You know what I mean? Right?

Baruch Sadogursky:

There is something into it. Right? Kubernetes is very complex and it requires a lot of time to learn. And today, if you are the one that have the time on their hands because of the staying home order, maybe learning Kubernetes is a good idea. It’s time consuming, but obviously that’s not the bottleneck of implementing DevOps.

Jayne Groll:

No, it’s not. And speaking of Kubernetes, we’ve been declaring a topic of the month. So, April is value stream management month, but May is Enterprise Kubernetes month. I think you’re going to be joining us at SKILup Days to talk a little bit about Kubernetes as well, so I’m excited about that as well.

Baruch Sadogursky:

Yeah. Yeah. And by all means, if you have time, the Kubernetes certifications, various Kubernetes certifications are a great way to spend this time with something very, very useful, that looks like is not going away. That’s a very good skill to learn, but yeah, I mean, the problem is not there. The problem is not your Kubernetes. The problem is people that collaborate or don’t collaborate with each other good enough. And today, collaboration is more critical than ever because of the situations we are in.

Jayne Groll:

Absolutely. Absolutely. And it’s our hope that we can provide more examples of what good collaboration looks like. We can provide more capabilities, in terms of what good collaboration looks like. But I also think going back and kind of wrapping up on our topic of liquid software, I really do love the analogy of liquid because liquid flows.

Jayne Groll:

And at some point, maybe it’s liquid collaboration. That, as the software is flowing, so go the humans, and that the barriers that would stop that ability to be continuous. So, a continuous collaboration approach is really necessary. Maybe liquid humans are needed for liquid software. Right?

Baruch Sadogursky:

Yeah. Yeah. Absolutely. I mean, yeah, liquid software is as much about collaboration as any other DevOps movement, if you wish. And obviously, as I mentioned, to empower those pipelines, to build the trust that we try to build between the software vendors and the software consumers, people have to talk to each other and have to collaborate, because unless we can trust the pipelines, we cannot trust the software.

Baruch Sadogursky:

If we cannot trust the software, we won’t take those updates. And if we won’t take those updates, there is no liquid software, there are no frequent updates. Zoom will get bashed until they roll something out in two months from now and it’ll be too late for them. And the unemployment system won’t be updated in another 20 years, which will actually cost people their income, their money, eventually their lives.

Jayne Groll:

Yeah, no, I absolutely agree with you. And again, if we’ve learned any lesson from this, it’s that lead, follow or get out of the way. I mean, that’s really what it comes down to. That accepting status quo, particularly in societies, global societies that are accelerating faster than ever before, where the shelf life of human skill is shorter than it’s ever been before.

Jayne Groll:

Again, do you learn COBOL or do you learn Kubernetes? Well, COBOL is going to have a certain lifespan now. Kubernetes or whatever comes after Kubernetes is going to be longer. So the message, I think is continuous. Continuous updates, liquid software, continually updating humans. We update our software. We need to update our humans and I’ll coin the term liquid human. We call them hybrid humans at DevOps Institute, but I think maintaining a kind of liquid mentality is important.

Jayne Groll:

And I think that when we come out of this, and we will … Maybe it’ll be in short spurts, maybe it will be permanent, but when we will, that we don’t go back to the way things were. That we do look for a new normal.

Jayne Groll:

And there’s probably a few new normals ahead of us, but that the work that you’re advocating, the work of JFrog, the work of the DevOps community, hopefully the work of the DevOps Institute provide the service that the world needs in order to realize the benefits of the new normal. So thank you for that.

Baruch Sadogursky:

Yeah. I’m personally very optimistic about the new normal. As I mentioned, some of the things that we have to come up with as a replacement for those other things will keep being around after the old things that we miss now will come back, and this is great. We actually create more than we used to have, and this is what it’s always been about.

Jayne Groll:

Yeah, absolutely. Absolutely. So I’m going to invite you back in a few months, and then we’re going to see where we are in the evolution. And I don’t know when that is. I really don’t know when that is. I don’t know.

Jayne Groll:

They’re saying we’ll be open in May. I don’t know when that is. I don’t know when we start to really move forward to a new normal. But I would invite you back because I said, I’m hoping that these very hard lessons, that have been particularly hard on humans and on certain segments in the business world globally, are lessons that we can use.

Jayne Groll:

So that, heaven forbid this ever happens again or something that is as disruptive … Maybe, hopefully not another pandemic, but something that we can persevere through all of that.

Jayne Groll:

Baruch, thank you so much. As I said, May is Enterprise Kubernetes month. So, our virtual SKILup Day is going to be focused entirely on Enterprise Kubernetes. And I appreciate very much the fact that JFrog is going to be a sponsor of that.

Jayne Groll:

So I invite anyone listening to come up and register. It’s free. It is a conference experience, so we have a lot of interactive aspects of the event that the Humans of DevOps can take advantage of. And certainly visiting with our friends at JFrog will be one of them.

Baruch Sadogursky:

For sure. We are very excited and looking forward and obviously it will be great. So yeah, by all means, please go ahead and register. And one of the benefits of the whole move to virtual events, most of them are free. So, there’s definitely no reason not to attend.

Jayne Groll:

Absolutely. Thank you. So please send my personal regards to the team at JFrog and thank you for spending some time with us today.

Jayne Groll:

So again, I’m Jayne Groll of the DevOps Institute. You’ve been listening to the Humans of DevOps podcast with … I know I’m going to mess this up, so I’m so sorry … with Baruch Sagadorsky. Did I get it right?

Baruch Sadogursky:

It’s close.

Jayne Groll:

Oh, it’s close. Developer advocacy with JFrog. Wishing you well. Stay safe and looking forward to another episode. Thanks very much.

Baruch Sadogursky:

Thank you.

Outro:

Thanks for listening to this episode of the Humans of DevOps podcast. Don’t forget to join our global community to get access to even more great resources like this. Until next time, remember, you are part of something bigger than yourself. You belong.

Upskilling IT 2023 Report

Community at DevOps Institute

related posts

[EP112] Why an AIOps Certification is Something You Should Think About

[EP112] Why an AIOps Certification is Something You Should Think About

Join Eveline Oehrlich and Suresh GP for a discussion on Why an AIOps Certification is Something You Should Think About Transcript 00:00:02,939 → 00:00:05,819 Narrator: You're listening to the Humans of DevOps podcast, a 00:00:05,819 → 00:00:09,449 podcast focused on...

[Ep110] Open Source, Brew and Tea!

[Ep110] Open Source, Brew and Tea!

Join Eveline Oehrlich and Max Howell, CEO of tea.xyz and creator of Homebrew, to discuss open source including "the Nebraska problem," challenges, and more. Max Howell is the CEO of tea.xyz and creator of Homebrew. Brew was one of the largest open source projects of...