Tag: open source

  • Tasktop Delivers Viz Value Stream Management App

    Tasktop Delivers Viz Value Stream Management App

    Tasktop this week announced the general availability of an application to manage projects based on a flow framework that captures events as they occur across the software development toolchain.

    Company CEO Mik Kersten said Tasktop Viz is the first substantiation of the flow framework in software that provides a value stream management application to enable organizations to more effectively manage software development projects. Tasktop Viz doesn’t necessarily replace project management software as a system of record for billing purposes, for example, said Kersten; however, it does provide a way to visually model the software development environment in a way that makes it easier to identify bottlenecks and potential opportunities to reallocate resources.

    Kersten said from a management perspective, existing approaches to managing software development projects are broken. They depend too much on software developers to manually enter data to update a project management application, which they don’t do as often as they should because it takes time away from developing software, he said. Tasktop Viz takes advantage of application programming interfaces (APIs) to automatically capture metadata from across the software development toolchain, providing instant visibility into software development projects without requiring DevOps teams to engage in an additional coding, Kersten noted.

    Tasktop Viz

    Kersten has been making a case for the adoption of an approach to managing software development projects based on the flow framework for several years. Tasktop Viz now provides IT teams with an application that is designed from the ground up using that model.

    Of course, the concept of value stream management applications has been around for some time. The debate now is not so much over the value of these applications as much as it is which architectural approach to building these applications best reflects modern software development. As the number of dependencies between projects continues to expand, it’s becoming next to impossible for organizations to manage multiple software development projects relying solely on spreadsheets and existing project management applications.

    That issue is further exacerbated by the fact that businesses are now more dependent on software than ever, so any delay to a project could have a dramatic impact on future revenues.

    Kersten said that despite the obvious business benefits, the biggest obstacle to adoption is more psychological than technical. Many software development teams are reluctant to adopt tools that might illustrate how chaotic the software development process within their organization really is. It’s critical that organizations that adopt value stream management applications do so with an eye toward wanting organizations to improve rather than just deploying another tool to measure metrics that are used to increase the amount of code being written.

    Of course, there’s no correlation between the amount of code being written and the quality of the applications being developed. The real challenge is to not just accelerate the rate at which applications are being developed, but also to ensure that once those applications are deployed in a production environment the number of fixes required are kept to the absolute minimum possible.

    — Mike Vizard

  • DevOps Chats: Liquibase, Open Source and Databases, with Datical

    DevOps Chats: Liquibase, Open Source and Databases, with Datical

    Datical has been a leader in DevOps for databases since even before the term was used. What many may not realize was that the open source Liquibase project was a major contributor to Datical’s success. So it should be no surprise that in filling the open president’s seat, they would tap an open source veteran.

    Dion Cornett, has open source credibility from his time at Red Hat and MariaDB. In his new role, he is responsible for marketing, sales and product. He is determined to see the same kind of success with open source business models he has been part of before, succeed at Datical.

    In this discussion we hear from Dion on his plans and views.

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

    [soundcloud url=”https://api.soundcloud.com/tracks/771223528″ params=”color=#ff5500&auto_play=false&hide_related=false&show_comments=true&show_user=true&show_reposts=false&show_teaser=true” width=”100%” height=”166″ iframe=”true” /]

    Transcript

    Alan Shimel: Hey, everyone, this is Alan Shimel for devopsy.kinsta.cloud. You’re listening to another DevOps Chat. We’ve got a nice DevOps Chat for you today, happy to introduce you to the newly minted president at Datical, Dion Cornett. Dion, I didn’t mess that up, did I?

    Dion Cornett: You didn’t, and thank you very much for the introduction, and I’m very happy to be on your podcast today.

    Shimel: Thank you, Dion. So, listeners of our podcast are no strangers to Datical. You know CTO Rob Reeves is a frequent guest here, and you know, we make fun of him and his Big Lebowski sweaters, but it’s always refreshing to hear Robert’s take on DevOps and databases, the DevOps market in general, and pontificating on anything Robert likes to pontificate about.

    But it’s also nice to have a fresh face from Datical. So, first of all, welcome aboard.

    Cornett: Thank you.

    Shimel: And I always like to give my audience a flavor for who’s talking. Dion, would you mind—I’m not asking you to read your bio, but would you share a little bit of your background for the audience?

    Cornett: Yeah, you know, I think what is more pertinent to Datical, I’d say if I had to characterize myself as that, I’m an open source guy. I bleed open source, I believe it’s a better production model for software, you know, this combination of a meritocracy of ideas and transparency leads to more elegant code.

    And so, you know, my background has been one of sort of having a career that aligns with this philosophy. So, I spent nine years at Red Hat in a couple different executive roles, run North American non-named account sales, and then ran global strategic alliances. I moved from there to take over sales at MariaDB, an open source database company that was—

    Shimel: Sure, [Cross talk] well.

    Cornett: Yeah. [Laughter] It wasn’t quite so popular in 2013, but we literally grew that business sevenfold in three years. That set me up for a CEO job. I did do something outside open source, but it was sort of a big data challenge, a company called ReachForce, that we successfully sold to another company in this space called Leadspace.

    And then, I was actually taking some time off, and introduction through a board member, spent some time with Datical, and just sort of fell in love with the concept. So, you know, my work at Red Hat, which obviously covers the very foundations of sort of IT into all the challenges as it relates to data at MariaDB and really sort of seeing the struggles that go on in building and maintaining a big, powerful database and how central that is to virtually any application. And you look at some of the top banking or telecom type customers we have, you know, data is the core of some of the value that many of these companies create.

    And then, even at ReachForce, right, big data company in the martech space, data is just paramount. It’s all maintained in databases and trying to update, upgrade, you know, maintain those databases in the context of fast changing application requirements is a non-trivial exercise.

    And so, when a company came along that had this combination of the open source that’s sort of in my DNA, and then solving this challenge around database change management, it just seemed like a great fit.

    Shimel: Absolutely. So, you know, another thing let us get out of the way early on, Dion, is—in today’s world, when we appoint a president, you know, there was a time, as we were talking, CEO and president were one person. And then, from a governance point of view, that’s a bad model. We have, for the most part, split off CEO from president. And presidents have various responsibilities. Some are more of a COO, some are more of a CRO, you know, meaning responsible for all revenue kind of activities.

    Just, what—where is your, you know, what are you responsible for at Datical?

    Cornett: Yeah, so, formally, my responsibilities are sales, marketing and product. And sort of the genesis here, and I have a great relationship with Derek, the CEO of the company. We’ve been on the job for a month now and I think we’ve teamed well, even through my recruitment process and through that.

    But Derek has built an amazing culture here. And this is one of the things you never know when you start a new job is just sorta, you know, what is the synergies between the team, and Derek’s a very effective leader, he’s built a great culture. In terms of the depth and the value of the solution that Datical has, it’s sort of undisputed, right? We’ve got multiple high six figure customers and even a seven figure customer. I mean, these are solving demanding problems in big enterprises, and the company’s been very good at that.

    I was brought on for a very specific charter, right? How do we better engage and collaborate with our community to push innovation out to the market faster? And as your listeners may know, Alan, the real foundation for Datical is an open source project called Liquibase. Liquibase is—and one of the things I also did at Red Hat was, I ran the venture portfolio there. I’ve looked at literally hundreds of open source business plans, made a dozen plus investments and sat on a few boards of companies that Red Hat invested in. And Datical was sort of a top 10—or Liquibase was sort of a top 10 project as you look at the thousands and thousands of open source projects.

    And so, how do we work with that community, again, in a collaborative fashion that we’re really pulling innovation, pushing it back, and driving as much value into this challenging database problem as we can?

    Shimel: Absolutely. So, you said a lot, there. Let’s pick ‘em off one by one.

    So, first of all Liquibase. It’s something we’ve mentioned in past podcasts and in past content type webinars and stuff that we’ve done with Datical. But—and maybe you’re gonna change this, and I hope you will, Dion. I know in my mind, when I think of Datical, I don’t think Liquibase. And though Liquibase has a tremendous following, I just don’t—it seems like, and I’m not blaming anyone or anything like that, but I just, I don’t connect the two. And maybe that’s part of what your mission needs to be.

    Cornett: Absolutely. We want this connection to not only exist, but be robust. Again, we’ve got a great team here, great culture, great people. We wanna use that to be excellent stewards of the community.

    You know, part of what motivates people here is, this is a big challenge to solve. You sort of commented earlier about sort of the maverick kinda style that Robert has, and I think it’s properly so, right? We’re breaking some glass. We’re trying to create a new mold, a new direction for how and a very important development of how IT operates.

    And so, that’s Liquibase. And, you know, Datical is a set of capabilities built on top of that. We need to make it clear, though, that that core is our foundation, and that, you know, we wanna be, again, as good of stewards as we can of that Liquibase community, 15 million plus users, and you will see this become much more of our identity, in everything we do, how we act, behave, where we invest and what we do to try to push value forward into our customers.

    Shimel: Excellent. So, Dion, let me throw another thing along the open source stuff out at you. You know, as I spoke to you earlier, I have my own views of where the market is, you know, builders, the democratization of IT, as I call it.

    I also have a view of where the open source movement, you know, the whole open source community is. And that is, you know, we’ve moved from the cathedral and the bazaar to what I call the big brother stage of open source where, seemingly, every open source project was kinda sponsored or in fact owned by a single vendor. And that kind of inhibits collaboration on an industry basis, right? It’s like, Oracle wasn’t gonna contribute to MS SQL. I mean, MS SQL was far from open. [Laughter]

    But let’s say, once Oracle bought—

    Cornett: MySQL.

    Shimel: MySQL. Look, that was the reason Maria and a host of others popped up, right? Because people weren’t sure that Oracle was gonna play nice, frankly.

    But that’s old. The model is what I call the foundational era of open source, where we have, like, the Linux Foundation or the Eclipse Foundation or the Cloud Foundry Foundation. You know, they’re these not-for-profits that are, they quote-unquote own the open source project to the community’s benefit.

    I’m not aware of a foundation in, like, the Liquibase kind of playing field. Are you, or is that—

    Cornett: No, you know, and I’m very familiar with the foundation model. Yeah, obviously, Red Hat was sort of a pioneer in this with Fedora.

    Shimel: Sure, yeah.

    Cornett: And we actually implemented a foundation model at MariaDB as well. And so, even if you look today, MariaDB.org is a separate legal entity—

    Shimel: Yep.

    Cornett: Separate president and set of engineers. And the real purpose there is to provide confidence to the users that, regardless of vendor acquisitions as you discussed, that the capabilities of that project will go forward.

    And we don’t—again, I’m a month on the job here—we don’t have a foundation today with Liquibase.

    Shimel: Right, I’m not asking you to—don’t say anything you can’t [Cross talk]. [Laughter]

    Cornett: Yeah. [Laughter] It’s certainly a consideration. [Cross talk] But it’s, you know, it’s a measured one, right? So, what are the trade-offs, what are the value? And as you talked about the different phases of open source, I think it’s—it’s unfortunate that there was a point where open source was sort of a, you used the phrase democratization of technology; that was a phrase also used by Matthew Szulik, the former CEO of Red Hat.

    Shimel: Uh huh.

    Cornett: And he also talked about this fashion statement thing where people, open source is a success. I mean, everything you use on the Internet, IT today is open source for the most part. So, it’s clearly won, right? That battle’s sort of over, and people said, “Hey, I wanna glom onto this.”

    And so, there was sort of a group of companies that attempted to co-opt the branding around open source without really understanding the nuance of what it meant to be a steward of the community and collaborate well. And in some ways, I think that sort of provided a skewed vision. And we’ll avoid that at Datical. You know, again, we’re gonna be good stewards of the community. We recognize that it’s gotta be participatory. Most of the community members are people that have jobs at companies you would recognize in IBM or SAP or other places like that. And they’re taking time to be part of this community, because it provides some value to what they’re doing and their professional life.

    And so, we’ve gotta recognize that we have to contribute to that value, attempt to sort of usurp what they’ve done, and go off and monetize it without the, never work, because if it really is important functionality, someone will just do a fork and monetize that around.

    It’s really about providing these synergies—and I used this phrase earlier where you have a meritocracy of ideas and that sorta lifts everyone up, right? So, we can provide higher value to customers that need the specifics of an enterprise offering or as we have brands, as you say, Liquibase Pro. And then the people that are, they’re part of the community for their specific needs that they’re getting multiples of return also on that investment for the problems they’re trying to solve.

    And it really is sort of this, community is the perfect word for it, because it really is this joint effort to make everyone better off.

    Shimel: Great. Dion, we’re running out of time. I wanted to give you one more thing to chew on. I want you, in your mind, to fast forward six months, 12 months, take your pick—what effect have you had on Datical?

    Cornett: You know, I think that we have to look at—so, part of what we’re doing is, we’re taking a blank sheet and we’re saying, hey, if we were gonna start today and look at the trends in the market, cloud, microservices, NoSQL, NewSQL, you know, these are all things that weren’t necessarily the case when Liquibase was originally founded by Nathan 10 years ago, what would it look like today versus where we are? And where are the gaps between the perfect DevOps solution for database and what we have in the community today, and sort of how do we help guide and steward that into a place that really solves the problem that we’re facing and we’re going to face as the IT infrastructure continues to evolve?

    And so, I think, if we can look back in a couple years and say, “Look, this has been the evolution of the product, and it’s much easier to use, it’s much easier to implement, the time to value is relatively quick,” and it’s really allowing, at the end of the day here, we’re trying to let database keep up with all of the DevOps advances that have been made around the application. If we see the database not being the drag on application, timeliness, speed, effectiveness that it can be today, then I know we’ve been successful in sort of this re-instillment, re-engagement in the community, and pushing that innovation into the product.

    Shimel: Yep. I agree with you. Dion, I promised I’d get you out of here in 15 minutes. I know you have another appointment. But I want to invite you back. Maybe we’ll do a video next time and we can dig in a little deeper. And, you know, I love talking about open source, as do you, it appears, and our audience does, too.

    Cornett: [Laughter]

    Shimel: So, I think it’s a great topic for us to continue. I wanted to wish you all the success in the world, though, with Datical. It’s a great bunch of folks over there, and you’re a great addition to the team.

    Cornett: Oh, thank you very much. I appreciate the kind words, Alan. And yeah, I look forward to future conversations. It’d be great.

    Shimel: It certainly well. Alright, Dion Cornett, new president at Datical, joining us here on DevOps Chat. This is Alan Shimel, and you’ve just listened to another DevOps Chat. Have a good day, everyone.

    — Alan Shimel

  • Top Six Open Source Tools for Monitoring Kubernetes and Docker

    Top Six Open Source Tools for Monitoring Kubernetes and Docker

    Kubernetes and Docker are two of the most commonly heard buzzwords in modern DevOps conversations. Docker is a tool that enables you to containerize and run your applications, and Kubernetes provides you with a platform to orchestrate or manage these containers—since managing thousands of containers manually using the Docker CLI would be a practical nightmare. 

    However, simply running thousands of containers and managing them through Kubernetes is not enough. You have to observe and analyze them properly to ensure that your services are up and running optimally. This process is called Site Reliability Engineering (SRE), a term initiated and popularized by Google. Observability and analysis are a major element of SRE. It can be subdivided into the following three areas:

    • Monitoring allows you to extract numerical metrics from your applications and resources that can later be visualized and analyzed to present the current state of your resources. Once the metrics are extracted, they can be used to set up alert rules, facilitate profiling and debugging, and make better SRE decisions.
    • Logging enables developers to debug their containers in case of a failure. Containers are short-lived, and so are logs. Kubernetes and Docker do provide a native way of browsing container logs, but it is very limited in its functionality. As a result, a centralized log plane is a must in any container-oriented environment.
    • Tracing allows you to debug services running on a network and follow a request trail until the source of a problem can be determined. In a microservices architecture, when multiple services/containers send requests to each other to perform one business task, a proper tracing solution is necessary. 

    This article will list and describe the most effective open source tools for monitoring and analyzing services running as containers. 

    Prometheus

    Prometheus is the first tool that comes to mind when discussing open source monitoring solutions. It is very popular in the development community, and it is a graduate of the CNCF project, which focuses on the monitoring side of SRE. Originally created and open sourced by SoundCloud, Prometheus simplifies the process of pulling numerical metrics based on a time series from a given metrics endpoint. It was built to monitor highly dynamic container environments.

    Prometheus is subdivided into three elements: Prometheus server, Alertmanager and exporters. Exporters are stand-alone, independent processes/containers that can be run on your target resource to generate and export metrics via a metrics API. The Prometheus server then takes care of service discovery as well as pulling metrics from the exporters to be stored in the Prometheus DB and later used for visualization or alerting. Alertmanager is responsible for setting up alert rules, analyzing the data in the Prometheus DB and sending alert messages to multiple receivers if a certain rule is triggered. There are a huge number of exporters, listed here, which are both officially supported by Prometheus and maintained by the community. 

    Prometheus has become the industry standard for monitoring cloud-native architectures. While it is known for its simplicity with service discovery, its ease of use, its alerting capacity and its integration with Kubernetes, its polling architecture is not ideal. Currently, the metrics endpoint has to be reachable by the Prometheus server. However, there is a pushgateway implemented in Prometheus that enables metric pushing rather than polling. Another downside of Prometheus is that it doesn’t scale well. This problem can be fixed in the Thanos adaptation of Prometheus.

    Related tools and technologies: Grafana, Cortex, Thanos, Prometheus Exporters, Alert Manager, Istio, Prometheus Operator.

    Figure 1: Prometheus Graphs. Source: SoundCloudDev.

    Grafana

    Grafana is an open source metric analytics and visualization suite. It allows you to create custom dashboards using data taken from multiple sources, such as Prometheus, Elasticsearch, MySQL, Postgres and Redis. Furthermore, Grafana has its own alerting system and a role-based access control (RBAC) system for software. As a data visualization tool, Grafana is famous among Prometheus users for enabling them to effectively visualize the metrics stored in Prometheus. There’s a wide array of official and community-built custom dashboards for various data sources in Grafana, allowing users to easily set up dashboards (found at this link) and get on with monitoring. Grafana offers another associated product called Loki, which aggregates logs in Kubernetes and integrates well with the Grafana UI.

    Related tools and technologies: Loki, Prometheus.

    Figure 2: Kubernetes Resources Dashboard in Grafana. Source: Replex.

    Elastic Stack

    Elastic Stack is a group of open source products from Elastic designed to help users search, analyze, and visualize data from any type of source, in any format, in real time. This product was previously recognized as the ELK stack, where each letter of the acronym stood for one of the company’s main products: Elasticsearch, Logstash and Kibana. Elastic Stack provides both monitoring and logging solutions with the help of its big data database, Elasticsearch. 

    In order to aggregate logs, people tend to use Elasticsearch for storage, Logstash or Fluentd for log streaming, and Kibana for visualization. Fluentd is not part of the Elastic Stack, but it is widely used with Kubernetes to stream logs in place of Logstash, the tool provided by Elastic Stack. Similarly, Metricbeat is used to scrape metrics and set up visualizations on Kibana. The enterprise version of Elastic Stack ships with X-Pack, a set of additional tools that enables reporting, alerting and role-based access control (RBAC), among other features. By default, the Elastic Stack GUI, Kibana, does not support RBAC. You have to use the previously mentioned enterprise version of Elastic Stack to enable it.

    Related tools and technologies: X-pack, Metricbeat, Logstash, Kibana.

    Figure 3: Discover View in Elasticsearch. Source: XpresServers.

    Sensu Go

    Sensu Go is a telemetry and service health checking solution for multi-cloud monitoring at scale. It allows you to gain visibility into servers, containers, services, applications, functions and connected devices across any public or private cloud. Sensu can be run side-by-side with Prometheus to get the best out of both solutions, or it can be run natively without Prometheus. Sensu works best alongside Prometheus, since exporting application level metrics to Prometheus requires loading the Prometheus SDK to your application codebase and exposing a metrics endpoint. The endpoint is then scraped and stored in the Prometheus Server. This may sound like a lot of work—and, sometimes, it is. Sensu avoids this complexity by using a sidecar concept. A Sensu agent is deployed alongside your application. The agent constantly collects and exposes metrics to the Prometheus server without your having to change the application codebase.

    Sensu works without Prometheus too. It can operate natively in Kubernetes, where it has its own server to store and visualize metric data exposed by the previously mentioned Sensu agent.

    Related tools and technologies: Prometheus.

    Figure 4: Events in Sensu. Source: becon.

    Sysdig Inspect

    Sysdig has two open source products: Sysdig Inspect and Falco. Here, we’ll focus on Inspect, which monitors and captures the container processes running in your system and allows you to drill into those processes for post-incident forensics. This enables you to analyze your application performance, troubleshoot errors and monitor any processors that could be misbehaving. Furthermore, if your system is breached, Sysdig allows you to understand how the breach occurred and what data was taken in the process. Sysdig Inspect is a very powerful tool that focuses on the performance tuning and security investigation of your system. 

    Related tools and technologies: Grafana, Sysdig, Sysdig Falco.

    Figure 5: Sysdig Inspect Overview. Source: Sysdig.

    Jaeger

    Jaeger is an end-to-end distributed tracing solution that was open sourced by Uber Engineering. It allows you to monitor and troubleshoot transactions in complex distributed systems. In modern microservice architectures, the majority of operational problems fall under the umbrellas of networking and observability. When a service failure occurs, you have no idea how the requests have gone from service to service over the network to complete a single business transaction. This makes debugging extremely difficult. Currently incubating under CNCF, Jaeger uses tracing to enable root cause analysis, performance and latency optimization, and distributed transaction monitoring. Jaeger works out of the box with Istio, a popular service mesh implementation open sourced by Google. 

    Related tools and technologies: Prometheus, Jaeger, Zipkin, Istio.

    Figure 6: Jaeger Find Traces. Source: Jaeger.

    Conclusion

    These tools are widely used in the tech industry, and they all have their benefits. Most of these solutions, however, require skilled implementation and ongoing manual maintenance that can be a burden for DevOps teams and a distraction from the business. There’s no one solution that can cater to all of your requirements, since each tool focuses on one or two specific aspects of observability and analysis. By mixing these tools together, you can derive a unique solution for your individual business needs. 

    For ease of comparison, the chart below outlines the features offered by each of the tools discussed in this article.

    — Ran Ribenzaft

  • Challenges with the Rapid Adoption of Cloud-Native Technologies

    Challenges with the Rapid Adoption of Cloud-Native Technologies

    To keep pace with the ever-changing digital landscape, today’s enterprises are trying to move to the cloud as fast as possible. As more and more companies move toward implementing cloud native and open source technologies to enable agility and increased time-to-value, the operational impacts to the business can be significant. While these challenges can cause serious setbacks and complexity, if an organization addresses them from the very beginning as part of its cloud strategy, it can overcome any trial.

    The following are top challenges enterprises may face when rapidly adopting cloud native technologies. 

    Intense Competition and Market Pressure

    As businesses fight to keep their competitive edge, they are feeling significant pressure to adopt new initiatives to achieve greater business velocity. New initiatives such as data analytics, data science and cloud computing are being leveraged by organizations to gain an advantage while disrupting how businesses have traditionally been performed. 

    Lack of Maturity and Interoperability of Many Open Source Technologies

    Understanding the cloud landscape is a critical component for kicking off any digital transformation project, but the state of each technology varies significantly. Because so many of these technologies are open source, getting a true understanding of maturity can be challenging. Some solutions are in very early stages with low commercial adoption and support, while other solutions are at a later stage and much more mature.

    Unfortunately, most of these technologies were not built to ensure interoperability from the beginning and the efforts of jamming together multiple disparate solutions from the cloud native landscape is complex and rife with challenges. This is why it is imperative that modern organizations ensure its solutions are enterprise-grade, ready to scale and nearly fool-proof to challenges. 

    No Clear Best Practices or Architectural Standards

    Because many of these new technologies lack maturity, the right information and proper documentation is scattered and difficult to distill. How can developers trust if an application is safe to include in their implementation if they are unsure whether documentation is correct and up-to-date? Since many of these open source technologies do not have well-defined best practices or architectural standards, implementing these solutions can be a huge business risk.

    Limited Choices for Leveraging External Domain Expertise and Support

    In order to ensure that organizations’  Day 2 operations go smoothly, they need the right expertise to provide guidance and support. Unfortunately, because so many of these technologies are new, there is a very limited pool of talent—both internally and externally—organizations can pull from. Organizations need the right technical support and services to operate and scale effectively, and with limited options available, the complexities of long term success can be a huge barrier.

    The Complexity of Managing Ongoing Operations

    Implementation is just the beginning, what really enables businesses to be successful with any digital transformation initiative is tackling ongoing Day 2 operations. This means ensuring that technologies work well together, the team understands how to leverage the technologies and organizations have an operational readiness plan for continued development and ongoing improvement. So many businesses focus their attention on what happens before production and as a result, success is not always achieved. 

    When beginning a cloud native journey, enterprises must be conscious of the inherent challenges rapid adoption poses. Competition, lack of maturity and best practices, limited choices and complexity can all complicate an enterprise’s journey to the cloud. Awareness of potential problems can help stop issues before they arise while easing the transition to the cloud.

    — Tobi Knaup

  • Cybersecurity Fears May Drive Shift to Managed DevOps

    Cybersecurity Fears May Drive Shift to Managed DevOps

    Should organizations consider using a managed service for DevOps to keep their platforms up to date and secure?

    The recent disclosure of a vulnerability that would allow open source Jenkins continuous integration/continuous delivery (CI/CD) platforms to be employed to launch a distributed denial of service (DDoS) attack highlights how challenging it is to secure the platforms on which many organizations now depend to build their most critical applications.

    The CVE-2020-2100 bug theoretically would have allowed cybercriminals to employ the Jenkins UDP discovery protocol to bounce traffic between servers until they could no longer respond. That same flaw also could be employed to launch DDoS amplification attacks against platforms connected to the internet. Those attacks can’t be stopped unless one of the servers is rebooted or its Jenkins service is restarted. The specific vulnerability discovered was fixed last month in Jenkins v2.219. IT organizations can either upgrade their Jenkins servers, disable the UDP discovery protocol or block the UDP port 33848.

    Tracy Miranda, director of open source community for CloudBees and member of the governing board for the Continuous Delivery Foundation (CDF), which oversees the development of Jenkins, said the bug itself is at best of medium severity. However, now that it’s been disclosed, the race is on to patch Jenkins servers or block UDP port 33848 before cybercriminals exploit the vulnerability on any public-facing instance of a Jenkins server.

    Given the fact that most IT organizations may not have resources at hand to patch their Jenkins servers quickly, Miranda said these and other potential future cybersecurity issues are a testament to why more organizations should rely on instances of Jenkins that are managed by third-party providers on their behalf. Organizations are spinning up more Jenkins servers than ever as they move to accelerate application development, and by relying on a managed service provider (MSP) to manage Jenkins, IT teams can focus more of their efforts on building and deploying applications rather than on managing CI/CD platforms.

    These days, more organizations are looking at DevSecOps as a best practice. Much of that focus, however, is on securing the applications that DevOps teams create; not nearly as much attention is being paid to securing the underlying platforms on which those applications are being built and deployed.

    It’s too early to say to what degree cybersecurity concerns might push IT organizations toward managed DevOps platforms. However, the more that cybersecurity teams participate in the DevOps process, the more they will ask questions about the fundamental security of the underlying platform.

    Of course, many IT teams often view managed services provided by third-party vendors as a threat to their existence. Nevertheless, DevOps platforms are among the most complex and expensive platforms to manage. In fact, that complexity is one of the main reasons so many organizations have hired site reliability engineers rather than relied on traditional IT administrators to manage DevOps platforms such as Jenkins. The issue that often comes up is just how many SREs an organization needs to hire before the cost of a managed DevOps platform becomes economically more appealing.

    — Mike Vizard

  • Red Hat Updates OpenStack Platform

    Red Hat today made available an update to its distribution of OpenStack that enables it to run on Red Hat Enterprise Linux (RHEL) 8 operating system.

    Sean Cohen, senior manager for product management at Red Hat, said Red Hat OpenStack Platform 16 combines three releases of OpenStack, known as “Rocky,” “Stein” and “Train” into a single update that the company plans to update on a streaming basis. In addition, future updates to Red Hat OpenStack Platform will be aligned with updates to RHEL 8, said Cohen.

    Red Hat OpenStack Platform 16 is also now being made available under a long-term support subscription as well.

    Red Hat Open Hybrid CloudCohen said Red Hat continues to see continued growth in terms of customers adopting OpenStack to build their own cloud computing platforms. Telecommunications companies have already made major investments in cloud frameworks that are now being extended to the network edge. Earlier this week, Orange Egypt announced it will be using Red Hat OpenStack Platform to deploy virtual network functions (VNFs) to manage network traffic. Enterprise IT organizations, meanwhile, are building their own private clouds on top of an open source platform.

    In some cases, enterprise IT organizations are relying on OpenStack to provide the foundation on which cloud-native applications based on Kubernetes are deployed, added Cohen. The goal is to take advantage of the virtual machines on which OpenStack is deployed to ensure isolation between instances of containerized applications running on top of Kubernetes. In other cases, IT organizations are taking advantage of tools such as Ironic to automate the deployment of Kubernetes cluster on bare metal machines, said Cohen.

    The relationship between OpenStack and Kubernetes, however, remains complicated because there are those organizations that are deploying Kubernetes clusters in on-premises IT environments without relying on any legacy infrastructure-as-a-service (IaaS) platforms such as OpenStack or VMware vSphere. Nevertheless, Cohen said Red Hat expects to see wider adoption of the Red Hat OpenShift application development and deployment platform, which is based on Kubernetes, on top of Red Hat OpenStack Platform 16.

    Cohen said there is a high correlation between organizations that have embraced best DevOps practices and usage of OpenStack as IT organizations move to automate the management of IT infrastructure at scale. The challenge organizations face is determining which modules of a very comprehensive OpenStack framework to employ. In fact, that challenge is why organizations rely on Red Hat to curate OpenStack on their behalf. In terms of adoption, Red Hat has the largest market share of any OpenStack distribution, said Cohen.

    It’s not clear to what degree the shift toward cloud-native applications based on Kubernetes might result in increased adoption of OpenStack. As organizations look to build hybrid cloud computing environments, the ability to deploy a common framework on-premises and in the public cloud has a certain appeal. However, public cloud computing platforms such as Amazon Web Services (AWS) already make available a proprietary platform for managing their IaaS environment. Many IT teams would naturally prefer to integrate with that AWS platform rather than replace it. Regardless of the approach to hybrid cloud computing, however, chances are high OpenStack will be in the mix in one form or another.

    — Mike Vizard

  • Collaboration Over Competition: How Companies Benefit from Open Innovation

    Collaboration Over Competition: How Companies Benefit from Open Innovation

    More and more technology companies are adopting open innovation initiatives. This is largely due to the realization of the benefits of working with outside experts to gain external perspectives and insights. This situation wherein an organization thinks beyond its internal resources for innovation and collaborates with external resources is known as open innovation. Open innovation is an opportunity for the company to utilize those external ideas and use them to develop innovative products and services. It may seem simple, but there is more to the collaboration process than just brainstorming.

    Open innovation is not easy, but the benefits are well worth the effort. Large technology organizations can establish a competitive advantage by partnering with promising startups. The startups benefit from this partnership by gaining access to venture capital resources, networking opportunities and the opportunity to reach a broader customer base, typically tied to a large company.

    Larger companies that typically change at a glacial pace can learn from their small yet evolving partners. Startup founders who are new to an industry can form valuable relationships and gain mentors with seasoned veterans.

    Overcoming Obstacles

    Technology organizations, large and small, should be prepared to make some changes in order to benefit from an open innovation environment. Indoctrinating an attitude of information sharing can be difficult in companies that have historically kept their intellectual property under wraps, and in many companies, even developing a collaborative environment across various departments can be an uphill battle.

    Breaking down barriers and overcoming this predisposition toward secrets is the only way to cultivate a collaborative environment. Ultimately, this effort begins in the boardroom. Once a large company has executive buy-in, the idea will begin to trickle from the top down, and over time it will take hold.

    The following steps will serve as important guidelines to tackle obstacles and sustain this synergy.

    Create Ecosystems Around Intellectual Property

    It is common for both parties to feel the need to protect their intellectual property, but that closed mindset defeats the purpose of an open innovation initiative. To overcome this obstacle, it is critical to emphasize the greater value and benefits that come from an open innovation ecosystem. All parties must realize that sharing their intellectual property into the ecosystem can accelerate the pace of innovation.

    Several companies have demonstrated the value of this approach. Tesla, for instance, made the decision to open source its patents a few years ago, recognizing that the company couldn’t make electric cars fast enough to slow the global carbon crisis. Elon Musk also pointed out that patents represented a small protection indeed against a determined competitor, and he thought the decision to open source would strengthen the company’s position.

    Assess Your Partners

    Adding a collaborator to your open network should not be a decision that is made lightly. The organization should be patient and cautious when selecting a potential partner. Weigh each company’s skills and capabilities and consider any disadvantages as well. Just because a partner is not perfect does not mean it is not the best choice, but knowing its weaknesses will help you address them before they become a problem.

    An organization should be assessing several prospective partners. Establish processes that ensure each potential collaborator is examined fairly, but critically so that the end goal to identify the companies that will not be a good fit is achieved. This practice is not just for your benefit alone—an unproductive collaboration can incur in heavy costs and wasted time to all parties involved.

    Adjust Business Models as Necessary

    Business models are designed to maximize benefits and minimize risks, but they should allow for some flexibility. In the real world, changes arise frequently, and business models must adapt. For instance, many partnerships established around open innovation initially began between two companies. Over time, it is a reasonable expectation to anticipate other players be added to the mix.

    A multi-dimensional ecosystem demands a different business model than a version that is two-dimensional because it may involve a different set of risks. It also has a more complex intellectual property pipeline where contributions and evolutions are more difficult to track and manage. As long as each partner is prepared to adjust business models, these changes can be effectively addressed.

    Foster a Collaborative Culture

    For the vast majority of companies, open innovation is not an organic process. Even within an organization, some departments tend to be closed off about their projects or even sharing information outside the company. An attitude of open innovation comes from internal change, and people are naturally resistant to change. To encourage a culture-shift, start from the top down.

    Support from leadership is critical for developing a culture that is compatible with open innovation. It is also important to identify the soft skills that a team needs to accomplish its goals. Open innovation itself is a skill that can be taught—problems arise when company executives expect it to be a trait that employees are inherently born with.

    There are numerous benefits of open innovation. Technology companies can produce more ideas in a shorter span of time mutually benefiting all businesses involved. There is an opportunity to create markets that did not exist prior to the development of the initiative.

    Of course, obstacles are also inevitable and more so because sharing valuable intellectual property outside the confines of a research and development department is always looked down upon. People, and business models, are naturally resistant to the significant change that comes with the open innovation approach. Whether a company can overcome these obstacles and see the many advantages is a matter of determination and a desire to adopt company-wide changes so that there is innovation across the organization.

    — K.R. Sanjiv

  • Wind River Launches CD Platform for Embedded Systems

    Wind River Launches CD Platform for Embedded Systems

    Wind River is fostering the adoption of DevOps processes for embedded systems by making available a continuous delivery (CD) platform that its customers can download.

    Glenn Seiler, vice president for open source strategy for Wind River, said the CD platform is based on an open source instance of the Jenkins continuous integration/continuous delivery (CI/CD) platform. The company envisions organizations downloading the platform as source code that Wind River will update regularly for use on top of the Wind River Linux platform, he said.

    Wind River is taking advantage of containers, Kubernetes, the Puppet IT automation framework and a repository to ease deployment of its CD platform, added Seiler. The company is committed to providing updates to that platform, which Wind River uses internally, at least every three weeks, said Seiler.

    Wind River expects most organizations will already have some form of a CI process in place. The CD platform provided by Wind River will focus on accelerating the delivery of applications built using whatever CI process organizations have in place today, said Seiler.

    In the embedded systems space, Seiler said the adoption of best DevOps practices is still in its very early days. However, with the rise of edge computing and the internet of things (IoT), organizations will need to find a way to securely deploy and update highly distributed application environments. To achieve that goal, many organizations will apply the same DevOps principles enterprise companies have been applying to their embedded systems, he said.

    To ensure the integrity of the software supply chain on which the Wind River CD platform is delivered, Wind River has also been certified under the ISO 9001:2015 specification for management processes, as well as the OpenChain specification for open source compliance.

    Wind River also has the Wind River Workbench for Linux LTS, based on the Eclipse integrated development environment (IDE), along with various system analysis tools including a trace visualization tool from Percepio AB to identify the root cause of anomalies.

    While it’s taken more than 10 years to drive the adoption of best DevOps practices in the enterprise, adoption in embedded systems environments should be faster now that the journey has begun. Organizations of all sizes are pushing more application workloads to the edge at faster rates. Traditional approaches to building and deploying applications on standalone embedded systems will not be able to keep pace. Organizations that build applications for embedded systems will be challenged by first acquiring the right tools and then adapting their internal culture to the complexities that arise when they need to build and deploy applications faster.

    Wind River is making the case for a curated open source CD platform that should enable organizations to spend less time on managing the underlying DevOps platform in favor of focusing more time and energy on changing their internal culture.

    — Mike Vizard

  • IBM Doubles Down on Cloud, But Is Krishna an Interim CEO?

    IBM Doubles Down on Cloud, But Is Krishna an Interim CEO?

    Big news from Big Blue yesterday: IBM CEO Ginni Rometty is stepping down after eight years at the helm. In her stead, IBM is doubling down on the cloud by appointing Arvind Krishna as CEO and member of the board. The real move to watch, though, is the appointment of Jim Whitehurst, Red Hat CEO, as president of IBM and a board member. It’s the first time that someone has held the title of president singularly, and it may portend where the real power will be at Big Blue (and Red).

    Krishna was previously head of the cloud and cognitive software division at IBM. He is an IBM veteran, having been with the company since 1990. His appointment sends a strong signal about how important the cloud is to IBM’s future. In some ways, it is similar to Satya Nadella’s appointment at Microsoft. But perhaps it’s not as strong a signal about the importance of the cloud than the $33 billion IBM spent acquiring Red Hat.

    It is that $33 billion investment and the importance of Red Hat to the new IBM that leads me to believe Krishna may, in fact, be a transitional CEO. The eventual appointment of Whitehurst as CEO and the rise of Red Hat leadership within IBM may in some ways be a reverse-merger type of thing, with Red Hat eventually running the show.

    IBM is a huge beast. A company of Red Hat’s size will find it difficult to change the culture and overcome the inertia at IBM. But frankly, IBM has been transitioning for almost all of Ginni’s time there. While the Enterprise Systems group (Z systems mainframe) is still the cash cow, bringing in a large percentage of revenue, IBM has increasingly tried to be a major cloud player.

    The SoftLayer acquisition didn’t quite give the company what it needed, as it was not a public cloud play. It left IBM unable to compete with Google, Microsoft Azure and AWS in true cloud services. After a few fits and starts, IBM decided that Red Hat, with its open source pedigree and cloud strength in both public and private cloud settings, could be the engine that could as it pertains to cloud.

    The street has not been kind to IBM over the last few years as a result of the botched cloud play, either. At a time when its tech competitors and the market in general have soared, IBM has actually fallen, as has its revenue.

    Mitchell Ashley, founder and managing analyst at Accelerated Strategies Group, agrees with this assessment. “We are seeing a sea change in big tech CEO strategy. Like Microsoft’s elevation of cloud exec Satya Nadella to CEO, IBM is rebuilding its executive ranks from the open source cloud software world with the appointment of Red Hat’s Jim Whitehurst as president and board member,” he said. “How soon the Krishna regime including old-line IBM executives gives way to Whitehurst and other Red Hat executives remains to be seen.”

    IBM CEO Change Elevates Open Source from Digital Anarchist on Vimeo.

    Ashley was not the only analyst to weigh in.

    “IBM bet the farm on cloud with the $34 billion Red Hat acquisition,” said Steve Duplessie, founder and senior analyst at Enterprise Strategy Group (ESG). “It was only a matter of time for the succession. Now the two top dogs running IBM are cloud purists. The old IBM died a while ago and they had to change—I like the move! Their stock lost 26% while the S&P rose over 150%. This lets them remake themselves before it’s too late.”

    Changes at the top don’t always portend success. There is a lot of hard work ahead at IBM as the company continues this journey. Competing with AWS, Google, Microsoft and other players in the cloud space is going to be challenging, given the head starts these players have already. But you really can’t count out IBM, ever.

    One other thing I wanted to mention: With IBM’s appointment of Krishna at IBM, there now are CEOs of Indian descent at Alphabet (Google), Microsoft and IBM. This is very positive for all persons of not only Indian heritage but for all people of color and nationalities. Kudos to these companies and these men for their success. Now, how about some women of color in these roles?

    — Alan Shimel

  • Atlassian Extends Road Mapping in Jira

    Atlassian Extends Road Mapping in Jira

    Atlassian has updated its Jira project management software for managing software development initiatives to add more capabilities to road mapping tools that teams use to share plans and update progress.

    Matt Ryall, head of product for Jira Software at Atlassian, said the road mapping tools first provided in Jira a year ago are now being employed by 45% of customers, making it the most rapidly adopted feature in 18 years the project management application has been around. Thus far, Atlassian reports end users have created more than 2 million road map items within Jira.

    Additional capabilities added to the road mapping tools provided in the latest edition of Jira include a drag-and-drop dependency mapping feature, support for filters to track projects more easily, a progress bar that indicates macro-level progress of each road map item, an ability to set up hierarchy levels on the road map, and the ability to set up road maps inline with a single click.

    road mapping

    In addition, a live Jira road map can now be integrated with Confluence, a complementary tool from Atlassian for managing projects spanning multiple teams.

    Atlassian currently offers several road-mapping solutions, including a team-level option in Jira, a program-level road map known as Portfolio for Jira and an enterprise organization-level version dubbed Jira Align. Atlassian has committed to making all three of these offerings work better together in 2020.

    In addition, Atlassian is committing to bringing together the various road-mapping capabilities is provides across its portfolio. The company plans to enable organizations to aggregate multiple road maps into a single consumable artifact and is exploring ways to make road-mapping functionality accessible beyond Jira users. As more organizations in the age of digital business realize how much their future success depends on timely delivery of software, the number of end users who want to be able to monitor the pace of application development projects continues to expand.

    In general, Ryall said Atlassian is focused on modernizing the end user experience for a project management application that is employed by more than 65,000 customers to manage software development projects. Much of that effort is focused on adding a visualization layer that will make it easier to navigate software development projects that continue to become more complex as organizations embrace, for example, microservices-based architectures.

    Competition across the project management category is intensifying as startups look to drive alternative approaches that they claim are designed to support, for example, a mobile-first user experience that makes it easy to visualize data. It’s not clear to what degree any of those offerings are gaining traction, but a significant amount of venture capital is starting to pour into the sector.

    In the meantime, visualization tools that promise to make it easier to appreciate what project events will have the most impact on the business are always going to be welcome. The real challenge, of course, remains getting developers to share accurate assessments of their progress with project team leaders. The good news is, it’s becoming much easier to automatically gather that data from the various repositories developers now rely on to check code in and out of various systems without having to ask anyone for an actual update.

    — Mike Vizard