Internet of Things Brings On Development Demands That DevOps Manages, Say Experts

1
Internet of Things Brings On Development Demands
That DevOps Manages, Say Experts
Transcript of a BriefingsDirect discussion on how continuous processes around
development and deployment of applications impact and benefit the Internet of Things
trend.
Listen to the podcast. Find it on iTunes. Get the mobile app. Watch for
Free: DevOps, Catalyst of the Agile Enterprise. Sponsor: Hewlett Packard
Enterprise.
Dana Gardner: Hello, and welcome to the next edition of the HP Discover Podcast
Series. I'm Dana Gardner, Principal Analyst at Interarbor Solutions, your host and
moderator for this ongoing discussion on IT innovation and how it’s making an impact on
people’s lives.
Our next DevOps thought leadership discussion explores how continuous
processes around the development and deployment of applications are both
impacted by -- and a benefit to -- the Internet of Things (IoT) trend. [Watch
for Free: DevOps, Catalyst of the Agile Enterprise.]
To help better understand the relationship between DevOps and a plethora
of new end-devices and data please welcome Gary Gruver, consultant, author and a
former IT executive who has led many large-scale IT transformation projects. Welcome,
Gary.
Gary Gruver: Thank you. It’s nice to be here.
Gardner: We're also here with John Jeremiah, Technology Evangelist at Hewlett
Packard Enterprise (HPE). He's on Twitter at @j_jeremiah. Welcome, John.
Learn how DevOps solutions unify development and operations
To accelerate business innovation
John Jeremiah: Hi, Dana. Thanks.
Gardner: Let’s talk about how the DevOps trend extends not to just traditional enterprise
IT and software applications, but to a much larger set of applications -- those in the
embedded space, mobile, and end-devices of all sorts. Gary, why is DevOps even more
important when you have so many different moving parts as we expect in IoT?
Gruver: In software development, everybody needs to be more productive. Software is
no longer just on websites and in IT departments. It’s going on everywhere in the
industry. It’s gone to every product in every place, and being able to differentiate your
product with software is becoming more and more important to everybody.
Gardner

2
Gardner: John, from your perspective, is there a sense that DevOps is more impactful,
more powerful when we apply it to IoT?
Jeremiah: The reality is it that IoT is moving as fast as mobile is -- and even faster. If
you don’t have the ability to change your software to evolve -- to iterate as there is new
business innovation -- you're not going to be able to keep up to be competitive. So IoT is
going to require a DevOps approach in order to be successful.
Gardner: In the past, we've had a separate development organization and approach to
embedded devices. Do we need to still to do that, or can we combine traditional
enterprise software with DevOps and apply the same systems architecture and
technologies to all sorts of development?
Software principles
Gruver: The principles of being able to keep your code base more "releasable," to work
under a prioritized backlog, to work through the process of adding
automated testing, and frequent feedback to the developers so that they get
better at it -- this all applies.
Therefore, for embedded systems you are going to need to develop
simulators and emulators for automated testing. A simulator is a
representation of the final product that can be run on a server. As much as
possible, you want to be able to create a simulator that represents the
software characteristics of the final product. You can then use this and trust it to find
defects, because the amount of automated testing you are going to need to be running to
transform your businesses is huge. If you don’t have an affordable place like a server
farm to run that, it just doesn’t work. [Watch for Free: DevOps, Catalyst of the Agile
Enterprise.]
If you have custom ASICs in the product, you're also going to need to create an emulator
to test the low-level firmware interacting with the ASIC. This is similar to the simulator,
but also includes the custom ASIC and electronics from the final product. I see way too
many organizations that are embedded and are trying to transform their process giving up
on using simulators and emulators because they're not finding the defects that they want
to. Yet they haven’t invested in making them robust so they can be effective.
One of first things I talk about to people that have embedded systems is that you’re not
going to be successful transforming your business until you create simulators and
emulators that you can trust as a test environment to find defects.
Gardner: How about working as developers and testers with more of an operations
mentality?
Gruver

3
Gruver: At HPE and HP, we were running 15,000 hours of testing on the code base
every day. When it was manual, we couldn’t do that and we really couldn’t transform our
business until we fundamentally put that level of automated testing in place.
For laser printer testing, there's no way we would have been able to have enough paper to
run that many hours of testing, and we would have worn out printers. There weren’t
enough trees in Idaho to make enough paper to do that testing on the final product.
Therefore, we needed to create a test farm of simulators and emulators to drive testing
upstream as much as possible to get rapid feedback to our developers.
Gardner: Tell us how DevOps helped in the firmware project for HP printers, and how
that illustrates where DevOps and embedded development come together?
No new features
Gruver: I had an opportunity to take over leading the LaserJet FW for our organization
several years ago. It had been the bottleneck for the organization for two decades. We
couldn’t add a new product or plans without checking the firmware, and we had given up
asking for new features.
Then, when 2008 hit, and we were forced to cut our spending, as a of lot of people out in
the industry at that time. We could no longer invest to spend our way out of problems. So
we had to engineer our solution.
Discover how to use big data platforms
To unlock value of Internet of Things
We were fundamentally looking for anything that we could do to improve productivity.
We went on a journey of what I would call applying Agile and DevOps principles at
scale, as opposed to trying to scale small teams in the organization. We went through this
process of continually trying to improve with a group of 400-800 engineers and working
through that process. At the end of three years, firmware was no longer the bottleneck.
We had gone from five percent of our capacity going to innovation to 40 percent and we
were supporting 1.5 times more products. So we took something that was a bottleneck for
the business, completely unleashed that capability, and fundamentally transformed the
business.
The details are captured in my first book, A Practical Approach to Large-Scale Agile
Development. It’s available at all your finest bookstores. [Also see Gary's newest book,
Leading the Transformation: Applying Agile and DevOps Principles at Scale.]
Gardner: And how does this provide a harbinger of things to come? What you’ve done
with firmware at HP and Laser Printers several years ago, how does that paint a picture of
how DevOps can be powerful and beneficial in the larger IoT environment?

4
Gruver: Well, IoT is going to move so fast that nobody knows exactly what they need
and what the capabilities are. It's the ability to move fast. At HP and HPE, we went 2-3
times faster than we ever thought possible. What you're seeing in DevOps is that the
unicorns of the world are showing that software development can go much faster than
anybody ever thought was possible before.
That’s going to be much more important as you're trying to understand how this market
evolves, what capabilities customers want, and where they want them in IoT. The
companies that can move fast and respond to the feedback from the customers are going
to be the ones that win. [Watch for Free: DevOps, Catalyst of the Agile Enterprise.]
Gardner: John, we've seen sort of a dip in the complexity around mobile devices in
particular when people consolidated around iOS and Android after having hit many
targets, at least for a software platform, in the past. That may have given people a sense
of solace or complacency that they can develop mobile applications rapidly.
But we are now getting, to Gary's point, to a place where we don't really know what sort
of endpoints we're going to be dealing with. We're looking at automated cars, houses,
drones, appliances, and even sensors within our own bodies.
What are some of the core principles we need to keep in mind to allow for the rapid and
continuous development processes for IoT to improve, but without stumbling again as we
hit complexity when it comes to new targets?
New technologies
Jeremiah: One of the first things that you're going to have to do is embrace service
virtualization and strategies in order to quickly virtualize new technologies and to be able
to quickly simulate those technologies when they come to life. We don't know exactly
what they're going to be, but we have to be able to embrace that and to bring
that into our process and methodology.
And as Gary was talking about earlier, the strategies of going fast that apply
in firmware, apply in the enterprise as well about building automated
testing, failing as fast as you can, and learning as you go. As we see
complexity increase, the real key is going to be able to harness that, and use
virtualization as strategy to move that forward.
Gardner: Any other metrics of success? How do we know we're succeeding with
DevOps? We talked about speed. We talked about testing early and often. How do you
know you're doing this well? For organizations that want to have a good way to
demonstrate success, how do they present that?
Gruver: I wouldn't just start off by trying to do DevOps. If you're going to transform
your software development processes, the only reason you would go through that much
turmoil is because your current development processes aren't meeting the needs of your
Jeremiah

5
business. Start off with how your current development processes aren't meeting your
business needs.
The executives are in a best position to clarify exactly this gap and get the organization
going down a continuous improvement process to improve the development and delivery
processes.
Most organizations will quickly find that DevOps has some key tools in the toolbox that
they want to start using immediately to start take some inefficiencies out of the
development process.
But don't go off to do DevOps and measure how well you did it. We're all business
executives. We run businesses, we manage businesses, and we need to focus on what the
business is trying to achieve and just use the tools that will best help that.
Gardner: Where do we go next? DevOps has become a fairly popular concept now. It's
getting a lot of attention. People understand that it can have a very positive impact, but
getting it in place isn't always easy. There are a lot of different spinning variables --
culture, organization, management. In an enterprise that's looking to expand in the
internet of things, perhaps they're not doing that level of development and deployment.
They probably have been a bit more focused on enterprise applications, rather than
devices and embedded. How do you start up that capability and do it well within a
software development organization? Let's look at moving from traditional development
to the IoT development. What should we be keeping in mind?
Gruver: There are two approaches. One is, if you have loosely coupled architectures like
most unicorns do, then you can empower the teams, add some operational members, and
let figure it out. Most large enterprise organizations have more tightly coupled
architectures that require large numbers of people working together to develop and
deliver things together. I don't think those transformations are going to be effective until
you find inspired executives who are willing to lead the transformation and work through
the process.
Successful transformations
I've led a couple of successful transformations. If you look at examples from the DevOps
Enterprise Summit that Gene Kim led, the common thing that you saw in most of those is
that the organizations that were making progress had an executive that was leading the
charge, rallying the troops, and making that happen. It requires coordinating work across
a large number of teams, and you need somebody who can look across the value chain
and muster the resources to make the technical and the cultural changes. [Read a recent
interview with Kim on DevOps and security.]
Where a lot of my passion lies now, and the reason I wrote my second book is, that I
don't think there are a lot of resources for the executives to learn how to transform large
organizations. So I tried to capture everything that I knew about how best to do that.

6
My second book, Leading the Transformation: Applying Agile and DevOps Principles at
Scale, is a resource that enables people to go faster in the organization. I think that’s the
next key launch point -- getting the executives engaged to lead that change. That’s going
to be the key to getting the adoption going much better. [Watch for Free: DevOps,
Catalyst of the Agile Enterprise.]
Gardner: John, what about skills? It’s one thing to get the top-down buy-in, and it’s one
thing to recognize the need for transformation and put in some of the organizational
building blocks. But ultimately you need to be have the right people with the right skills.
Any thoughts about how IoT will create demand for a certain set of skills and how well
we're in a position to train and find those people?
Jeremiah: IoT requires people to embrace skills and understand much broader than their
narrow silo. They'll need to develop an expertise in what they do, but they have to have
the relationships. They have to have the ability to work across the organization to learn.
One of the skills is constantly learning as they go. As Gary mentioned earlier, it’s not a
"done" for DevOps. It’s a journey of learning. It’s a journey of growing and getting
better.
Skills such as understanding process and understanding how things are working so you
can continuously improve them is a skill that a lot of times people don’t bring to the
table. They know their piece, but they don’t often think about the bigger picture. So it’s a
set of skills. It’s beyond a single technology. It's understanding that that they are really
not in IT -- they're really a part of the business. I love the way Gary said that earlier, and I
agree with him. Seeing themselves as part of the business is a different mindset that they
have to have as they go to work.
Then, as they apply their skills, they're focusing on how they deliver business value.
That’s really the change.
Gardner: How do you do DevOps effectively when you're outsourcing a good part of
your development? You may need to do that to find the skills.
For embedded systems, for example, you might look to an outside shop that has special
experience in that particular area, but you may still want to get DevOps. How does that
work?
Gruver: I think DevOps is key to making outsourcing work, especially if you have
different vendors that you're outsourcing to because it forces coordination of the work on
a frequent basis. Continuous integration, automated testing, and continuous deployment
are the forcing functions that align the organization with working code across the system.
When you're enabling people to go off and work on separate branches and separate issues
and you have an integration cycle late in the process, that’s where you get the dysfunction

7
-- with a bunch of different organizations coming together with stuff that doesn’t work. If
you force that to happen on a daily, or multiple times a day, basis, you get that system
aligned and working well before people spend time and energy working on something
that either don’t work together or won’t work well in production. [Watch for Free:
DevOps, Catalyst of the Agile Enterprise.]
Gardner: We have been exploring how continuous processes around development and
deployment of applications impact and benefit the Internet of Things trend. I'd like to
thank our guests, Gary Gruver, consultant, author and a former IT executive who has led
many large-scale IT transformation projects, and John Jeremiah, Technology Evangelist
at Hewlett Packard Enterprise on Twitter at @j_jeremiah.
Learn how DevOps solutions unify development and operations
To accelerate business innovation
And I'd also like to extend a big thank you to our audience for joining us for this DevOps
and Internet of Things innovation discussion. I'm Dana Gardner, Principal Analyst at
Interarbor Solutions, your host for this ongoing series of HPE-sponsored discussions.
Thanks again for listening, and come back next time.
Listen to the podcast. Find it on iTunes. Get the mobile app. Watch for
Free: DevOps, Catalyst of the Agile Enterprise. Sponsor: Hewlett Packard
Enterprise.
Transcript of a BriefingsDirect discussion on how continuous processes around
development and deployment of applications impact and benefit the Internet of Things
trend. Copyright Interarbor Solutions, LLC, 2005-2015. All rights reserved.
You may also be interested in:
• Big data generates new insights into what’s happening in the world's tropical
ecosystems
• DevOps and security, a match made in heaven
• How Sprint employs orchestration and automation to bring IT into DevOps
readiness
• How fast analytics changes the game and expands the market for big data value
• How HTC centralizes storage management to gain visibility and IT disaster
avoidance
• Big data, risk, and predictive analysis drive use of cloud-based ITSM, says panel
• Rolta AdvizeX experts on hastening big data analytics in healthcare and retail
• The future of business intelligence as a service with GoodData and HP Vertica
• Enterprises opting for converged infrastructure as stepping stone to hybrid cloud
• Redcentric orchestrates networks-intensive merger using advanced configuration
management database
• HP pursues big data opportunity with updated products, services, developer
program

8
• How eCommerce sites harvest big data across multiple clouds
• How Localytics uses big data to improve mobile app development and marketing

Internet of Things Brings On Development Demands That DevOps Manages, Say Experts

More Related Content

What's hot (20)

Similar to Internet of Things Brings On Development Demands That DevOps Manages, Say Experts (20)

Recently uploaded (20)

Internet of Things Brings On Development Demands That DevOps Manages, Say Experts