devopsy.kinsta.cloud

Tag: cloud platform

  • Banks’ Ability to Maximize Their Use of New Digital Technologies May Be Hindered by Slow Transition to the Cloud, New Research from Accenture Shows

    NEW YORK; Nov. 5, 2018 – Although the majority of banks recognize the advantages of building new information technology (IT) systems on cloud platforms, many have been slow to transition to the cloud, and only one in four has a solid strategy for cloud adoption, according to a new report from Accenture (NYSE: ACN).

    The report — titled “Cloud and Clear, Accenture Cloud Readiness Report – Banking” and based on a global survey of banking executives — finds that cloud adoption is integral to enabling banks to quickly add new online services, develop applications and improve customer experience. The top benefits that surveyed bankers expect from cloud adoption are the ability to scale their IT operations up — and down — as needed (cited by 77 percent) and the ability to launch applications faster (63 percent); only 49 percent cited cost savings as a top benefit.

    However, more than two-fifths (43 percent) of the bank executives surveyed said their banks lack a strategy for cloud adoption or have only started to implement basic cloud practices, and only 26 percent have what Accenture considers a solid cloud-adoption strategy, which includes ongoing measurement of efficiencies for continuous improvement.

    “To drive growth, banks must provide customers with a stellar digital experience, and cloud is a core component to accomplishing this,” said Andrew Poppleton, head of the technology advisory practice within Accenture’s Financial Services operating group. “Cloud provides the capacity and agility that banks need to use big data and analytics to identify the best products and services for their customers and to quickly develop and get them to market, which is crucial to keeping pace with competitors in this digital age.”

    According to the report, banks’ initial focus on the cloud has traditionally centered around infrastructure-related functions. The survey found that the two most-common functions that banks have moved, or are in the process of moving, to the cloud are disaster recovery (cited by 80 percent of respondents) and data backup (74 percent). Banks have also made progress in identifying which legacy applications are good candidates for early migration to the cloud, with the top three being retail customer access channels, such as email, mobile, internet and ATM (cited by 65 percent of respondents); foreign exchange (60 percent); and regulatory reporting (40 percent).

    However, the report identifies two impediments to banks’ wide-spread cloud implementation:

    • Existing legacy IT infrastructure. Nearly all bank executives surveyed said they expect that transitioning to the cloud will require new IT operating models. In fact, more than two-thirds (69 percent) intend to operate in a “bimodal” way — maintaining key legacy systems and those not easily replicated on cloud platforms, while transferring other systems and adding new applications in the cloud.
    • A skills gap. Although nearly nine out of 10 bankers (89 percent) surveyed rated the skills and experience of their employees in charge of cloud strategy and management as “mature” or “advanced,” they were less confident about their teams’ ability to manage cloud infrastructure and applications. Specifically, more than half (55 percent) rated their team’s skill level as “novice” or having “no experience” in managing cloud infrastructure, and 40 percent gave the same skill rating for managing cloud applications.

    “The cloud provides extraordinary opportunities for banks to transform how they conceptualize, develop, manage and sell their products and services — but this can’t be accomplished without the right talent in place to advance the cloud agenda,” said Chad Duncan, a managing director in Accenture’s Financial Services technology advisory practice. “At the same time, our research shows that many banks still lack a comprehensive strategy to rapidly shift to the cloud, leading to limited deployment. Because cloud is fundamental to how banks will conduct business and grow in the future, it’s time to move past the experimental phase and put an aggressive cloud transition strategy in place.”

    Methodology
    Accenture Research surveyed 35 banking executives responsible for technology and information security at global retail banks in North America, South America, Europe, Africa and Asia Pacific. The survey was conducted during November and December 2017.

  • Morpheus Data Extends DevOps Reach

    Morpheus Data Extends DevOps Reach

    Morpheus Data signaled it will pursue a more open strategy when it comes to enabling DevOps teams to manage multiple cloud platforms.

    Brad Parks, vice president of marketing and business development for Morpheus Data, said updates to the company’s namesake platform will make it possible for DevOps teams to swap in third-party modules and components even when Morpheus has already provided that equivalent functionality in the platform. This more open approach is being driven by a recognition that in many cases DevOps teams are being required to implement a specific corporate standard, Parks said.

    The latest edition of the Morpheus cloud management platform adds support for Kubernetes container orchestration software, Terraform configuration management software developed by HashiCorp and Microsoft Azure Resource Manager. Morpheus already has its own JSON/YAML scripting framework. But it’s clear more declarative approaches to managing cloud resources using, for example, Kubernetes or Terraform are starting to gain traction, Parks noted.

    At the same time, Morpheus is adding connectors to cloud services from Oracle and T-Systems.

    At its core, the Morpheus cloud management platform enables DevOps teams to consolidate analytics, reporting, governance, monitoring, logging and data protection within a single platform. Most organizations today either already are or expect to soon employ multiple clouds. Each of those clouds will be running different classes of workloads as IT organizations move to keep their cloud service provider options open. At the same time, Morpheus Data is betting that, rather than mastering separate management consoles for each cloud, organizations will move to centralize the management of multiple clouds as much as possible, Parks said.

    Parks noted Morpheus Data is also starting to see DevOps teams hooking up their preferred continuous integration/continuous deployment (CI/CD) platform to the Morpheus cloud platform. That approach provides them with a platform to manage applications starting the minute a CI/CD pipeline finishes pushing them into production, he said.

    CI/CD platforms and other DevOps tools are great when it comes to building and deploying applications, but currently a gap exists between how those applications are built and deployed versus managed when deployed in a production environment. Morpheus is designed to provide an open flexible approach to closing that gap, Parks said.

    Interest in various approaches to automating the management of IT is clearly on the rise. Most organizations simply don’t have enough IT staff to manage large numbers of applications being deployed at scale. In fact, Parks said Morpheus Data is seeing a marked rise in the number of managed service providers focused on DevOps and cloud management services.

    Regardless of how the integration of DevOps and cloud management is achieved, it’s clear IT processes are being transformed. The day when silos within IT organizations implemented various levels of IT automation based on various capabilities of platforms is starting to give way to a more comprehensive approach. The real challenge now is figuring out how to achieve that goal with the least amount of reasonable disruption to the culture of the IT organization as possible.

    — Mike Vizard

  • DevOps Chat: ParkMyCloud Saves Wolfpack App Over 50 Percent on Cloud

    DevOps Chat: ParkMyCloud Saves Wolfpack App Over 50 Percent on Cloud

    In this DevOps Chat, find out how the developers of Wolfpack (wolfpack.run) are saving over 50 percent in cloud costs by using ParkMyCloud’s cloud management platform.

    Though the cloud offers many benefits to app developers and others, controlling the costs can sometimes elude organizations big and small. Cloud sprawl can quickly lead to large cloud bills. In this chat, we speak with Jonathan Chashper, CEO of Wolfpack, about how his company got a handle on the cloud and is spending less than 50 percent of what it was paying before.

    As usual, the streaming audio of our conversation is below, followed by the transcript of our conversation.

    Audio

    Transcript

    [su_pullquote align=”right”] [su_custom_gallery source=”media: 118815″ target=”blank” width=”200″ height=”200″]Chris Parlette
    Chris Parlette is the Director of Cloud Solutions at ParkMyCloud. Chris helps customers reduce their cloud waste and manage their hybrid infrastructures by drawing on his years of experience working at various software startups. From SaaS to on-prem, virtualization to cloud, monitoring tools to cloud management platforms, and small businesses to large enterprises, Chris has seen it all and loves helping drive improvements to IT management.[/su_pullquote]

    Alan Shimel: Hey, everyone, it’s Alan Shimel of devopsy.kinsta.cloud, and we’re here for another DevOps Chat. Today’s chat is actually a great behind-the-scenes look at ParkMyCloud and how an app provider named WolfPack is using Park My Cloud to really kinda get a handle on their cloud costs, their cloud usage and optimizing the cloud for their applications.

    That being said, let me introduce you to our guest for today’s podcast. Our first guest is the CEO and founder of WolfPack, and his name is Jonathan Chashper. Jonathan, welcome to DevOps Chat.

    Jonathan Chashper: Thank you, Alan. Thank you for having me.

    Shimel: Thank you. And then our second guest is from ParkMyCloud, it’s Chris Parlette, and Chris, I hope I don’t mess up the title, but it’s director of Cloud Solutions?

    Chris Parlette: Correct, yes. And thanks, Alan, for having us—appreciate being here.

    Shimel: Not a problem. So, guys, this is gonna be a fun one, I think. Jonathan, why don’t you give our audience a little background—what exactly is this WolfPack app?

    Chashper: Sure. So, WolfPack, we are—you can look at us as a connected vehicle technology focusing specifically on helping individuals move together. Think about how much fun it is to move as a group. Imagine you are, you know, maybe several families moving in a few cars. It’s a lot of fun, but there are two major problems—one, keeping everybody together, making sure people don’t get separated, don’t get lost. And the second one is actually communicating with each other while traveling and doing it safely.

    Now imagine you are now doing it on model cycles, which is the first niche we’re focusing on. So, we’re solving those two problems with WolfPack, and by the way, people can take a look at WolfPack.run-R-U-N-to see how it works. But generally, what we do is, we provide, with a turn-by-turn navigation, to every one of the people driving or riding, while providing a radar showing the location of everybody else.

    So, we actually make sure that everybody gets there without being killed, without being lost, without getting into a ditch, and making the experience much less stressful, and much more fun.

    Shimel: And that’s a good thing. But Jonathan, well, you know, WolfPack represents sort of this new—it’s not really new. I’ve seen them now for a couple years, but I call them these instant group type of applications where you can create a group on the fly and that group can have two people or 200 people or more, right? And this group then does something. In the case of WolfPack, you know, they’re on a run or a trip of some sort, and you have Google Maps built into it or underlying it and, you know, all of these things, and they’re great.

    The problem is, as you probably are well aware, is you never know from moment to moment, you know, what your load is gonna be, right? How many groups, how big a group, you know, what’s going on?

    And so, in resource planning, that can be a problem. Talk to me—

    Chashper: Correct.

    Shimel: —yep. Talk to us a little bit about what you’re doing with ParkMyCloud and how that’s helping you.

    Chashper: Mm-hmm. Yeah, so you’re absolutely correct. What you said at the beginning, it’s true. We can—our users can immediately generate groups, it can be, they can generate any number of groups. It can generate—any group can have any number of people. They can plan a ride which can be as short as 10 miles or as long as 1,000 miles, so you’re absolutely right.

    And so, when we’re building the system, obviously, it’s a cloud based system. We are on top of AWS. What’s important for us is that the development costs, being a startup, is being-you know, kept to the minimum. And so, what we did, we actually looked into ParkMyCloud, and with ParkMyCloud, we got to the point that we control our development costs very, very tightly. Generally speaking, when you’re using a server in the cloud, as long as you don’t turn it off, it still works and you’re still being billed for it. And if your development teams go home-say, they work 9 to 7 every day, they go home at 7. If the servers are off, you’re still paying for them.

    So, what we did with ParkMyCloud, we actually manage the up time of those servers, so we only pay for the time we’re actually using those servers for our development for our staging.

    The other good thing we’re having is that we actually have quite good visibility. True—a nice supporting feature, which I think is pretty new, where we can see very, very quickly, you know, when is our load high, what is actually being used, what is not being used, if anything needs to shut down. So, we can manage that very, very easily. And specifically, on the park management, it’s almost a file and forget. Once you set it up once, you never need to go back to it, and it just does it for you.

    Shimel: Got it. So, Jonathan, let me just be clear, I want to make sure we’re talking apples to apples. So, when you say, you know, you’re not using the server overnight or something like that, are you saying that, with ParkMyCloud, you can schedule your instances to close? So, in other words, I’m only gonna use this instance from 9 to 5, and at 5:00, retire that instance, close it out, don’t keep it on my billing account, and when I start up again tomorrow at 9, I’ll start a new instance with that same configuration and kind of pick up where I left off, or is it something else?

    Chashper: So, first, we need to understand that you can use ParkMyCloud on your non-production instances. Production instances obviously need to be up 24/7, so you don’t want to shut them down.

    Shimel: Yeah, you’re right—I should have said that. [Laughter] No one wants to shut down their production.

    Chashper: Yes. So, what I was referring to, I was talking about our development and staging nonproduction systems.

    Shimel: Mm-hmm.

    Chashper: Yeah, so that’s for that.

    Shimel: But are you actually—I guess what I’m asking, are you retiring those instances and firing up new ones the next day, or is it—

    Chashper: Oh, this is—yeah, so this is done automatically for us. That’s the beautiful thing about ParkMyCloud. We actually, once you set up the parking schedule, it will shut down the server, say, 5 p.m. today, and it will turn it on back tomorrow morning, say, 9 a.m. We don’t need to do anything. This is done for us automatically, and so you know, where the developers come back into the office, as far as they’re concerned, nothing changed, they just keep working.

    Shimel: That’s fantastic. Chris, I just want to confirm with you—those instances are actually, you know, it’s sort of like immutable infrastructure, right? Those instances are retired, done, and then re-fired up the next day?

    Parlette: Yeah, so what this is—and it’s not terminating the instances, it’s actually just stopping them and then turning the same ones back on. And the nice thing is about that, is that now, you know, as people are working on these servers, you know, if you haven’t fully gone with data, you know, not being stored on those servers, you know, as you’re working on these same servers over and over and over again, you know, we stop them and then start them right back up again. So, even if people like your finance department or your QA department maybe have some things, you know, spun up, we can actually turn them off to save on the CPU and memory costs, and then we just turn them back on again.

    Shimel: I got it. Okay, because I was gonna say, you know, I mean, the immutable infrastructure angle is great, but in this sort of circumstance, it’s not just a configuration of the server, it was the state of the data that you were working on, and how the heck would you re-create that every morning? [Laughter] But that explains it—got it.

    So, Jonathan, you know, so ParkMyCloud is making you guys a lot more efficient here in the cloud and your use of cloud resources. Have you been able to quantify a percentage savings or anything like that?

    Chashper: Oh, absolutely—most definitely. If I want to be, you know, conservative, I would probably say that we’re saving around 55 [percent]to 60 percent.

    Shimel: Really?

    Chashper: At least. Frankly, I think we probably save close to 70 percent. We’re really very tight. We really manage to do that in a very tight manner.

    Shimel: So, beyond just closing down and firing up the instances the next day, what are some of the other things you’re doing with ParkMyCloud?

    Chashper: Oh, so—well, for us, the first thing is reporting and notification. If something happens, we get notified, we can actually control what we get to be notified about. So, then we’ll get notification about maybe a server that needs to be shut down or something like that.

    Reporting is very important to us so you can see your usage, you can see what is up and what is down.

    Shimel: Got it.

    Chashper: One of the cool things we just started using is actually a mobile app. So, there is a mobile app on iOS for ParkMyCloud. People can download it at the iOS store. And it actually happened with me, I was at dinner, and we—my guys had to stay a little later, and we had to, what we call, snooze one of the parking, you know, schedules, and I just did that through the mobile app, I didn’t even have to go online.

    Shimel: Fantastic. Very cool. So, guys—look, 50 percent or better savings, all these other benefits is certainly nothing to sneeze at. I think anyone would jump at it.

    But, Chris, let’s dive in with you now a little bit, if we can. You know, we have an audience here of DevOps sort of folks from all over the world, right, who would love to save 50 percent on their cloud bill. How is ParkMyCloud working with these DevOps engineers and DevOps teams to realize those kinds of savings and efficiencies.

    Parlette: Sure, yeah. So, we work, you know, plug into a lot of DevOps processes and toolkits straightaway. I mean, that’s something that we do internally, and then, you know, we certainly have all of our customers, from large enterprises down to small startups, all doing this same kind of thing.

    And so, there’s a few different ways to do that. One of the big things about DevOps is automation, and so, we have a full policy engine so you can set up the policies to look at tags, to look at locations to different regions that these servers were in, naming schemes—all these different kind of factors, and you can automatically apply schedules and assign to teams and decide on different actions to take based on these policies for any new instances that you spun up. So, that’s one automated way.

    We also have a full API, so you could plug into DevOps tools like Atlassian, Bamboo or Jenkins so that these tools can kind of reach into our API, temporarily override schedules, change schedules—all this as a part of an automated thing.

    One of the other things about, you know, DevOps—not just automation, but also just thinking whole business, kind of a holistic approach to this thing—and so, we also have a full, a very easy to use UI. So, if your users are, maybe are a little less technical or you don’t want them to necessarily have direct access to a database console or Azure or Google consoles, this can kind of be sort of a proxy for that console so that other users can come in and manage their end servers. They don’t have to ask IT to turn things on for them. They can kinda manage their own servers and do all this, but you can decide who has access to what through this approach.

    So, there’s a couple different ways we plug into these teams and these processes so that users can get their best bang for their buck.

    Shimel: Got it, got it, got it. Now, so, Chris, when you talk about plugging into things like Bamboo or Jenkins, what you’re really—I mean, I just wanna make sure I’m clear. You can build sort of the ParkMyCloud—not sort of, but you can build the ParkMyCloud policies right into your CI/CD pipeline. Is that what you’re saying?

    Parlette: Yes, so if you do things like using tags, especially as instances are spun up or spun down, a lot of people say, you know, “Hey, I put a million tags on every instance.” And so, that’s what the ParkMyCloud policies can read off those tags and kinda decide on the actions based on those tags.

    And then, as far as things like—you know, specifically, if we were to talk about Jenkins or some other build processes, you know, you can have these environments built out and shut down. And then Jenkins, at the beginning of some sort of build process, can call into ParkMyCloud and say, “Hey, I need these servers up, and so I wanna override these schedules.” Spin ’em up, the servers are up, it does all it needs to do, and at the end, it just either shuts them down or, if you don’t shut them down, then basically, if you only did an override, let’s say, for two hours? At the end of the two hours, we’ll revert back to our original schedule, so that way you don’t have to remember to turn the lights out when you leave the room, right? So, we actually can shut things down for you.

    Shimel: Excellent. And Chris, give me an idea of—so, when you guys are engaged, I mean, with that kind of API, with that kind of integration into the DevOps toolset, you’re probably getting a decent view of how DevOps teams are using ParkMyCloud. And you’ve given us some examples, here.

    Are there any—I guess what I’m trying to ask, are there any special things that the DevOps people are doing that perhaps waterfall or other types of technologies are not doing? Is there commonality? Is it very easy to say, “Oh, these guys must be doing DevOps, because here’s, you know, what they’re doing?” Have you noticed any of that?

    Parlette: Yeah.

    Chashper: Alan, before Chris answers, maybe I can jump in and give a quick lead-in to this.

    Parlette: Sure.

    Chashper: Sorry, Chris—hope you don’t mind.

    Parlette: Sure.

    Chashper: Actually, one of the things I wanted to say is that we don’t have DevOps people at all, and one of the reasons we’re using ParkMyCloud is because we’re actually—we’re not DevOps people. We don’t know what needs to be done. We don’t speak, you know, Chef or any one of those guys.

    So—but the fact that we could actually use ParkMyCloud and then the user interface they provided us with, the web client, at least it did for us, went away, and everything is handled at that level. So, for us, that was a big, big move forward.

    Shimel: Yeah, no—and that actually is a great point, Jonathan. And Chris, I’m sorry we hijacked you here for a second.

    Parlette: No problem.

    Shimel: But the problem is, in today’s market, frankly, hiring people who have Chef or Puppet or Jenkins or, you know, DevOps skills sets is (a) hard to find, and (b) very expensive, right? And so, if you can deliver benefits like 50 percent savings on cloud usage or allowing some automation like this, it certainly is a lot easier. I don’t know of any engineering team that wouldn’t like that.

    But, you know, Chris, what do you see? I mean, are most of the people using the UI, or are they just you know, kinda going to APIs and stuff like that?

    Parlette: Yeah, and you know, as we get into larger customers, you know, it definitely is a mix of the UI and the API. The other change that we see, the mindset shift that we see from people is, especially as they’re moving from on-prem to cloud, a lot of people, when they’re used to—a lot of people, when they’re coming from a [Dog barks].

    Shimel: Hey, Jonathan, I think Chris might have shut off there for a second. Chris, you there?

    Parlette: Here we go. I’m back—I’m back. So, one of the things that, you know, as people are moving from on-prem to cloud, you get a lot of people who are used to leaving things on 24/7, because you know, on-prem, as things are left—you know, if you spin things up, it’s not an additional cost, really, to leave them on 24/7.

    And so, in this shift to the cloud, they have this shift of mindset where, if they’re being charged by the hour or it’s something that, you know, every hour that it’s left on is additional cost.

    Chashper: Yeah.

    Parlette: And so, when people are used to leaving things on 24/7, now the shift we have with people using ParkMyCloud, especially in an automated way, is to leave things off 24/7 and then kind of spin them up on demand. And so, that’s really where I think, you know, Jonathan and his team use that a lot, where things can be off, you know, turn on a schedule, but also just turn on on demand so, as the users have access to the UI and the API, they can turn things on on demand, as opposed to having to turn things off on demand.

    Shimel: Fantastic. Well, guys, you know, as I told you off mic, the time here goes really quick. [Laughter] We’ve already been on 18 minutes, beyond the 15 minutes I told you we would do, so we’re gonna need to wrap this one up right here. But first of all, Jonathan Chashper, CEO, Founder of WolfPack, it’s WolfPack.run is the URL, correct?

    Chashper: Yes, absolutely—thank you.

    Shimel: Thank you. Thank you for sharing with us today and giving us a little peek behind the kimono of how you guys are using ParkMyCloud. Continued success for any of our motorcycle riding friends or any group out there who—you know, I encourage you to check WolfPack out, it’s a pretty cool app.

    And Chris, thanks so much for sharing a little bit more, you know, a little behind-the-scenes look and a little bit more of a peek at, you know, ParkMyCloud and how it’s helping companies. Thanks for being on today’s DevOps Chat.

    Parlette: Yeah, Alan. Thanks for having me. I really appreciate it.

    Chashper: Alan—thank you very, very much. It was a pleasure.

    Shimel: Thank you, guys. You know what? We’re recording this right before Thanksgiving—have a happy holiday. Happy holidays to everyone listening! You’re probably listening to this after Thanksgiving, but it’s the start of the holiday season, so happy holidays, everyone.

    This is Alan Shimel for devopsy.kinsta.cloud and DevOps Chat. Hope to see you soon on another chat, and have a great day.

    — Alan Shimel

  • Salesforce Moves Compliance to the DevOps Left

    Salesforce Moves Compliance to the DevOps Left

    Compliance is one of those tasks that trips up developers multiple times over. More times than anyone cares to admit developers finish an application only to discover there are any number of compliance controls that were overlooked. Sometimes, bringing that application into compliance is trivial. Other times, it can delay the rollout of the application by months.

    To help mitigate this issue, Salesforce this week added a Heroku Shield module to Heroku Enterprise, its container-based platform-as-a-service (PaaS) environment available to IT organizations needing access to a dedicated platform running on Amazon Web Services (AWS). Dylan Steele, vice president of marketing for the Salesforce platform, says Heroku Shield enables developers to comply with a broad range of mandates ranging from payment card standards to any number of healthcare regulations. The goal is to automate meeting regulation compliance by making it possible for developers to automatically add the appropriate controls to their applications.

    Heroku Shield requires mandatory encryption for data at rest in addition enforcing a stricter level of TLS encryption for applications deployed in a Shield Private Space. It also enforces trusted IP ranges and manages application-level authentication tokens.

    The move is part of the effort to shift left more of the responsibility for IT security, by making it easier for developers to address security and compliance issues early in the application development process. That’s when the cost of fixing those issues is substantially less than when that application is deployed in a production environment.

    In general, there’s not a lot of love lost between developers and auditors. Developers often feel that the controls needed to achieve compliance are not all that clear. Auditors feel that many developers are simply too lazy to implement the appropriate controls in the first place. In truth, however, there’s a lot of commonality between the control required to meet various compliance requirements. That creates a lot of opportunity for PaaS environments that abstract the underlying infrastructure away from developers to automate implementing a broad range of compliance and security controls. The biggest issue now may be simply making developers aware that’s possible.

    In the meantime, DevOps teams that regularly struggle with compliance issues at the back end of an application release cycle might want to reconsider their approach. Whether it involves a PaaS environment or not, there’s a strong case to be made for moving testing of compliance controls as far left as possible in the application development process. Not only does that approach minimize the number of compliance issues that may arise before and after an application is deployed, but the documentation generated during the development process should substantially reduce the amount of time auditors spend crawling over and around an IT environment. That’s critical, since most auditors charge by the hour, and every minute they spend on an audit also increases the probability of finding something else that needs to be fixed.

    — Mike Vizard

  • DevOps Q&A: SUMO Heavy’s Brodie on the Hybrid Cloud

    DevOps Q&A: SUMO Heavy’s Brodie on the Hybrid Cloud

    Not every expert says DevOps will enable the hybrid cloud. One says it never will. Find out why in this Q&A with Bob Brodie, partner and CTO of SUMO Heavy, a digital commerce design and development firm.

    David Geer: How will DevOps enable the hybrid cloud?

    Brodie: I do not feel that DevOps enables the hybrid cloud. Hybrid clouds are very complex, and DevOps would need to be supported by a team of network and system engineers. The beauty of being on the cloud is that it’s API-driven. You can provision servers and configure and deploy with tools that are easy to set up.

    Bringing dedicated servers into that environment which do not have tooling to match the cloud counterpart means that it can’t be solely maintained by a DevOps team, but rather needs much more support. We have worked in hybrid cloud environments, and have experienced firsthand the amount of effort and resources required to create and maintain a hybrid cloud environment.

    Geer: What makes hybrid clouds more complex than being either fully in the cloud or fully dedicated?

    Brodie: When mixing dedicated servers with cloud servers on the same network, you will likely need parallel VPNs set up so you are not sending your data unencrypted over the internet and so that if there is a networking issue there is a fallback. Setting up full-duplex, parallel VPNs can require both expensive hardware and specialized resources to configure and maintain that hardware.

    Geer: Why have tools such as some of the ones AWS is building (OpsWorks On-Premises) not hit the inflection point where these are making it easier to enable the hybrid cloud? What other tools are examples of this?

    Brodie: OpsWorks On-Premises is just one tool that AWS has. They also have VPNs to connect your dedicated hardware to your virtual private cloud. These tools aren’t necessarily to make it easier to run a hybrid cloud, but they allow you to utilize existing AWS provisioning and deployment tools to set up and deploy to your existing infrastructure.

    Geer: Why do you think these tools never will make it easier to enable the hybrid cloud?

    Brodie: Hybrid cloud infrastructure is inherently more complex than being fully dedicated or fully cloud-based. They are two different types of ecosystems, and are managed differently. They require different skill sets and levels of experience within those ecosystems.

    Geer: Please expand on your following statement: “The problem is that the methodologies behind the two schools of thought (Separate sysops and dev versus DevOps) are vastly different in thought and practice, and to really bring dedicated hardware into a cloud-based architecture is very cumbersome, especially if the dedicated environment is not utilizing complex virtualization.” What does the difference in schools of thought have to do with it? What does bringing dedicated hardware in have to do with it?

    Brodie: My point here is that “DevOps” is often misused. It isn’t a tool, it is a mindset. It’s a combination of development and sysops. A “hybrid cloud” is a mix of dedicated hardware and cloud infrastructure. My main point here is that DevOps practices do not make managing enterprise developments easier across hybrid cloud environments. Hybrid cloud environments are inherently more complex than the alternatives (dedicated hardware or fully cloud-based).

    The reason for this is that development is becoming more and more intertwined with sysops. For example, five years ago you could download MAMP onto your Mac and be fully set up with a PHP & MySQL development environment. Modern PHP development requires a developer to understand specific versions of PHP and MySQL, how to manage them, how various applications communicate with each other and, often times, many other applications (Redis, MongoDB, Queueing systems, etc.). These can be complex to maintain, and even if they are distributed to developers from the ops team, it will often be in the form of Vagrant boxes or Docker files, which the developer will be required to understand.

    This is the nature of the “DevOps” school of thought, versus the traditional “SysOps separate from development” school of thought. The target environment needs to be considered when developing an application. For example, if the application will reside on AWS then the developer will need to have a solid understanding of the tools available on AWS (RDS, EC2, ElastiCache, etc.). The maintenance of these becomes part of the developer’s job, which was not present in previous times. Adding the extra complexity of a hybrid cloud means that the developer needs to understand the underlying network architecture of that system.

    Geer: In what way is the hybrid cloud a settling point, mostly built on marketing?

    Brodie: Hybrid cloud environments are not ideal. They are costly to maintain and provide more points of failure because of the networking between them. What it can do is provide a halfway point for CIOs that are reluctant to move to the cloud, so they can see some of the benefits on the cloud side and hopefully move their organization to the cloud.

    Geer: What are organizations afraid to let go of?

    Brodie: Many organizations are scared to move to the cloud. They may have a CIO/CTO that is afraid of “losing control.” I have seen many environments such as this.

  • The Maturing OpenStack Cloud: Just Do it

    The Maturing OpenStack Cloud: Just Do it

    We’ve been thinking about the maturing cloud a lot lately. In fact, we’re participating in a panel discussion on the topic at the upcoming OpenStack East in August in New York. As we’ve spoken with OpenStack users and vendors throughout the world we’re hearing a common theme about the maturation of OpenStack clouds—the results seldom look like what was originally planned. Disruptive technologies, changes in personnel, the evolution of business strategies … these are all unplanned events that can change what the mature cloud eventually will look like. The challenge for today’s DevOps teams is to how to maximize the opportunities for success, despite the likelihood of unanticipated changes. After all, as many DevOps folks know, between the design of an application or system and the completion of that project, they often are faced with a change in the initial business requirements of the project.

    Lao Tzu famously said, “The journey of a thousand miles begins with one step.” While that is true, we’d add that the journey will be much smoother if that one beginning step relies on careful consideration of the best shoes for the journey: shoes that are comfortable, durable and multi-purpose. Shoes that see you through any type of weather. The right shoes won’t necessarily get you to the right destination, but they certainly will make your journey easier and better ensure you complete your journey—even if you end up somewhere other than where you thought you would.

    So what do shoes have to do with the OpenStack cloud? We liken choosing the right shoes to making the right architectural decisions at the beginning of your OpenStack journey. One of the most important architectural decisions DevOps can make early in the development of the OpenStack cloud is to include a comprehensive data protection strategy, accompanied by the appropriate tools to facilitate that. There are three primary reasons to include data protection from the start:

    • It’s a “given” for VMWare and other enterprise applications for a reason. One might argue there’s no need for data protection, since everything lives in the cloud. But we all know files, applications and systems can fail for no apparent reason. It’s much easier to recover from a failure if you have a data protection strategy in place, rather than having to suddenly (and often quickly) re-create the lost resources.
    • It will make it easier to manage OpenStack and cloud resources. As your OpenStack cloud grows, you may need to move workloads to different availability zones or data centers. While you may have a general idea of your resources needs at the beginning of the DevOps process, chances are things will change along the way.
    • It makes for a smoother upgrade process. Chances are that you are not rolling out your OpenStack cloud with the exact OpenStack version you started with. For example, one of our customers started on Kilo and now is migrating to Mitaka for its rollout. Their migration is much easier because the customer knows everything is backed up, available to use and restored in its new, upgraded environment.

    Sure, the OpenStack and the cloud paradigm is enabling DevOps and business users to upend many “traditional” approaches to IT. But that doesn’t mean you should toss out every strategy that’s considered “traditional.” After all, if you were buying your shoes for your thousand-mile journey, would you only buy one pair? Or make sure you have a spare just in case?