The Cloudcast

Peter Mattis (@PeterMattis, Co-founder/CTO of @CockroachDB) talks about the evolution of scalable SQL databases, the challenges of globally scalable data management, how Kubernetes has evolved to manage stateful applications, and lessons learned running Kubernetes and CockroachDB. SHOW: 438SHOW SPONSOR LINKS:MongoDB Homepage - The most popular database for modern applicationsMongoDB Atlas - MongoDB-as-a-Service on AWS, Azure and GCPDatadog Homepage - Modern Monitoring and AnalyticsTry Datadog yourself by starting a free, 14-day trial today. Listeners of this podcast will also receive a free Datadog T-shirtCLOUD NEWS OF THE WEEK - http://bit.ly/cloudcast-cnotwSHOW NOTES:CockroachLabs HomepageCockroachDB - Scalable, survivable, SQL database OperatorHub HomepagePeter Mattis on the Google Kubernetes PodcastTopic 1 - Welcome to the show. Before we dive into our discussion, tell us a little bit about your background in working on scalable technologies. Topic 2 - Today we’re going to mash together a couple of popular (and complex) topics - the growing use-cases on Kubernetes, and the growing need to synchronize data for anywhere access. Let’s start with the data side of the equation - tell us about the basics of your creation, CockroachDB, and the challenges it solves.Topic 3 - What are some of the use-cases that are driving more scalable SQL usage vs. more traditional SQL database models?Topic 4 - When Kubernetes first got started, the focus was on scalable stateless (cloud-native) applications. How are you beginning to see the trend towards companies becoming more comfortable with stateful applications (e.g. databases) on Kubernetes?Topic 5 - One of the new technologies that’s making it easier to get databases onto Kubernetes is “Operators”. CockroachLabs has been one of the leading platforms supporting this technology. Can you talk a little bit about your experience with Operators and how it images the way Kubernetes teams (developers or platform teams) about databases on Kubernetes.Topic 6 - What are some of the lessons learned from deploying CockroachDB onto Kubernetes?FEEDBACK?Email: show at thecloudcast dot netTwitter: @thecloudcastnet

Sam Charrington (@samcharrington, Host of TWIML & AI Podcast) talks about AI & ML trends in 2020, frameworks to understand usage patterns, hot new technology to explore, how long projects take to succeed, and the inherent bias built into every AI & ML model.SHOW: 437SHOW SPONSOR LINKS:Datadog Homepage - Modern Monitoring and AnalyticsTry Datadog yourself by starting a free, 14-day trial today. Listeners of this podcast will also receive a free Datadog T-shirtMongoDB Homepage - The most popular database for modern applicationsMongoDB Atlas - MongoDB-as-a-Service on AWS, Azure and GCPCLOUD NEWS OF THE WEEK - http://bit.ly/cloudcast-cnotwSHOW NOTES:TWIML Homepage (Podcasts, eBooks, etc.)eBook: The Definitive Guide to  ML PlatformsStudy Groups & Education TWIML Conference HomepageSam Charrington on The Cloudcast in 2019 (Eps.321) Topic 1 - Welcome back to the show. Let’s start with the broad set of TWIML activities that you’re working on these days. Topic 2 - You focus on AI & ML every week, across a lot of different domains and usages. It’s a broad scope. If you had to focus it on Enterprise/Business leaders, how do you structure a conversation around how to align business opportunity and technology choices?  Topic 3 - What are some of the most commonly used technologies being deployed around AI/ML systems? Any big shifts over the last couple of years? Topic 4 - You’ve been around Cloud Computing and DevOps communities, which required companies to go through some people/process change to achieve success. What are the people/process changes that you typically see with AI/ML environments?Topic 5 - If somebody asked you how they can put a timeline on when they’ll see value around their AI/ML, is that a realistic ask? What are the factors that go into achieving success in AI/ML projects?Topic 6 - What are some of the interesting usages of AI/ML that you’ve seen in use recently?Topic 7 - There has been quite a bit of discussion recently about bias in AI/ML algorithms. Can you explain what this means and how it could impact the system’s decision making?FEEDBACK?Email: show at thecloudcast dot netTwitter: @thecloudcastnet

SHOW: 436DESCRIPTION: Stephen Pinkerton (@spnktn, Product Manager @datadoghq) talks about the latest trends that Datadog has observed from monitoring serverless applications in the public cloud, adoption rates, as well as how serverless is deployed along with containers. SHOW SPONSOR LINKS:MongoDB Homepage - The most popular database for modern applicationsMongoDB Atlas - MongoDB-as-a-Service on AWS, Azure and GCPDatadog Homepage - Modern Monitoring and AnalyticsTry Datadog yourself by starting a free, 14-day trial today. Listeners of this podcast will also receive a free Datadog T-shirtCLOUD NEWS OF THE WEEK - http://bit.ly/cloudcast-cnotwSHOW NOTES:Datadog's State of Serverless ReportDatadog's Monitoring Lambda GuideTopic 1 - Welcome to the show; Tell us about your background, as you’ve worked for a number of interesting companies prior to Datadog. Topic 2 - Datadog recently released a report about trends in serverless usage. Before we dive into some of the highlights and insights, tell us about the scope of the report. Are you able to tell anything about the types of companies being monitored? (industries, size of company, etc.)Topic 3 - Highlights from the State of Serverless reportThe majority of functions are still fairly small, and don’t run very long (small memory footprint). Node.js and Python are the leading languages/frameworks being used (almost 90%)Nearly half of EC2 users have adopted LambdaEarly adopters of Containers (ECS) have adopted Lambda - 80%+Discuss which event-driven data sources are the most frequently usedTopic 4 - As you’re analyzing this usage data, what are some of the things you’re thinking about as a Product Manager for functionality that you can provide to help better monitor these rapidly changing environments?FEEDBACK?Email: show at thecloudcast dot netTwitter: @thecloudcastnet

SHOW: 435DESCRIPTION: James Higginbotham (@launchany; Founder, Author, API Architect) talks about frameworks for evaluating API usage, developer perspectives on APIs, versioning APIs and some thoughts on new trends in API usage. SHOW SPONSOR LINKS:Datadog Homepage - Modern Monitoring and AnalyticsTry Datadog yourself by starting a free, 14-day trial today. Listeners of this podcast will also receive a free Datadog T-shirtUpCloud - World’s fastest cloud serversUpCloud (promo) - Sign up for free, receive a $50 credit and try us out!MongoDB Homepage - The most popular database for modern applicationsMongoDB Atlas - MongoDB-as-a-Service on AWS, Azure and GCP[DONUT RUN DONATIONS] CLOUD NEWS OF THE WEEK - http://bit.ly/cloudcast-cnotwSHOW NOTES:James on The Cloudcast Eps.135 (“Building Better APIs for Business Success)LaunchAny HomepageRecent keynote: “APIs, Microservices, and Serverless: The Shape of Things to Come”Topic 1 - Welcome back to the show; It’s been a while, let’s reintroduce you to our audience. Tell us about the types of things you work on.Topic 2 - There are so many things going on with APIs these days, sometimes it’s hard to know where to start. Do you have a framework that you use to help companies think about APIs?  Topic 3 - Are there different perspectives that developers have if they are dealing with APIs for monolithic applications vs. microservices applications vs. external APIs? Topic 4 - What is some of thinking around dynamically changing environments (e.g. DevOps, Agile) and APIs (versioning, testing changes, etc.)?Topic 5 - What are some of the more critical things that you’re always reinforcing and educating people about APIs? Topic 6 - We’re seeing more companies emerge that just deliver APIs as part of an ecosystem of services. How has this changed application development?FEEDBACK?Email: show at thecloudcast dot netTwitter: @thecloudcastnet

SHOW: 434DESCRIPTION: Christian Posta (@christianposta, Field CTO @soloio_inc) talks about the trends that are shaping the Service Mesh space, including emerging standards, application patterns, and interactions with API gateways. SHOW SPONSOR LINKS:UpCloud - World’s fastest cloud serversUpCloud (promo) - Sign up for free, receive a $50 credit and try us out!MongoDB Homepage - The most popular database for modern applicationsMongoDB Atlas - MongoDB-as-a-Service on AWS, Azure and GCPDatadog Homepage - Modern Monitoring and AnalyticsTry Datadog yourself by starting a free, 14-day trial today. Listeners of this podcast will also receive a free Datadog T-shirt[DONUT RUN DONATIONS] SHOW NOTES:Christian Posta’s blog (Service Mesh, Cloud-native apps content)Solo.io’s blog Christian on PodCTL podcast (discussing Istio)Topic 1 - Welcome to the show; you’ve been on PodCTL in the past. Tell us about your background, as you’ve been very active with application developers and distributed systems for quite a while.  Topic 2 - A few years ago, Service Mesh came onto the scene as a big deal (Istio, Linkerd, etc.) and people were trying to figure out what it was, what it did, etc. The technology has evolved quite a bit, but people are still oftentimes confused. How should we think about what a Service Mesh does (or doesn’t do)? Topic 3 - What are the most common use-cases when Service Mesh is being used? What are some of the places where Service Mesh is discussed, but probably shouldn’t be used? (API-Gateway, code in an application, etc.)Topic 4 - Sometimes we have a technology space that has lots of implementations (e.g. Kubernetes, Swarm, Mesos, etc.) that eventually converge into a single industry choice. But Service Mesh still has lots of implementations. Are they all really different? Will we see industry convergence around a standard? Do we need a standard?Topic 5 - What are some of the areas where you expect that we’ll see advancements in Service Mesh in 2020?Topic 6 -  What are some of the best ways for people to start either learning more about Service Mesh, or trying out the technology to see if it makes sense for them?FEEDBACK?Email: show at thecloudcast dot netTwitter: @thecloudcastnet

SHOW: 433DESCRIPTION: Gene Kim (@RealGeneKim, Author, DevOps Researcher) talks about his new book “The Unicorn Project”, the follow-on to “The Phoenix Project, and the newest trends shaping DevOps and business success. SHOW SPONSOR LINKS:MongoDB Homepage - The most popular database for modern applicationsMongoDB Atlas - MongoDB-as-a-Service on AWS, Azure and GCPDatadog Homepage - Modern Monitoring and AnalyticsTry Datadog yourself by starting a free, 14-day trial today. Listeners of this podcast will also receive a free Datadog T-shirtUpCloud - World’s fastest cloud serversUpCloud (promo) - Sign up for free, receive a $50 credit and try us out![DONUT RUN DONATIONS] SHOW NOTES:The Unicorn Project (homepage)The Unicorn Project (resource guide)The Phoenix Project on The Cloudcast (Eps.79)The DevOps HandbookDevOps Enterprise Summit (event)Topic 1 - Welcome back to the show. It’s hard to believe it’s been 7 years since you were last on the show. Before we get into the Unicorn Project, let’s talk briefly about some of the other things you’ve been doing since The Phoenix Project (DevOps Handbook, DevOps Enterprise Summit, etc.) Topic 2 - Let’s do a Cliff Notes version of The Unicorn Project. It picks up 2-3 years after The Phoenix Project - what are the challenges now facing Parts Unlimited?Topic 3 - Without giving away the ending, there is a sense that the Unicorn Project team essentially says that DevOps is dead and they are going to do things a new way. Is that the message of the story?Topic 4 - The hero of this story (Maxine) is what I’d call a “25x engineer”. She’s portrayed as super-elite in her engineering skills. I’ve also noticed that the State of DevOps report is also now adding a focus on these super-elite teams. Is that a good thing to focus on, since so few teams identify like that? Topic 5 - This book talks about 5 “ideals”. How would you stack rank them in terms of ease or difficulty in achieving?THE FIRST IDEAL: Locality and SimplicityTHE SECOND IDEAL: Focus, Flow, and JoyTHE THIRD IDEAL: Improvement of Daily WorkTHE FOURTH IDEAL: Psychological SafetyTHE FIFTH IDEAL: Customer Focus Topic 6 - In Unicorn Project, you have a very broad set of characters. Do you do this so that a broad set of people can identify within these challenging environments? FEEDBACK?Email: show at thecloudcast dot netTwitter: @thecloudcastnet

SHOW: 432DESCRIPTION: Aneel Lahkani (@aneel, Go To Market Consultant and Advisor) talks about how SaaS impacts today's cloud computing decisions, the challenges of running a profitable SaaS business, migrating to SaaS services, and how public clouds compete and partner with SaaS offerings. SHOW SPONSOR LINKS:Datadog Homepage - Modern Monitoring and AnalyticsTry Datadog yourself by starting a free, 14-day trial today. Listeners of this podcast will also receive a free Datadog T-shirtUpCloud - World’s fastest cloud serversUpCloud (promo) - Sign up for free, receive a $50 credit and try us out!MongoDB Homepage - The most popular database for modern applicationsMongoDB Atlas - MongoDB-as-a-Service on AWS, Azure and GCP[DONUT RUN DONATIONS] SHOW NOTES:Exploring the SaaS Business Model (Cloudcast #390 - March 2019)Product & Marketing for Engineers (Aneel’s Blog)Topic 1 - Welcome back to the show. For anyone that doesn’t know you, tell us a little bit about your background, and some of the areas you focus on now.Topic 2 - What is SaaS in 2020? (Delivery model, acquisition and usage model, experimentation model, etc.). Is it a finite thing, or has it been componentized? Topic 3 - Is it possible to build software and then make it SaaS, or do the business and software have to be built in parallel with the SaaS offering?Topic 4 - Does anyone actually “migrate” (successfully) existing types of applications to SaaS, or should the focus of SaaS be on breaking free from old approaches? Topic 5 - We’re seeing many “tools” companies become “platform” companies and offer things that blur the lines between SaaS and something else (GitHub, Salesforce, etc.). Does this become the way that they compete against AWS or Azure?Topic 6 - Why do you think AWS doesn’t have a broader set of SaaS offerings at this point? FEEDBACK?Email: show at thecloudcast dot netTwitter: @thecloudcastnet

SHOW: 431DESCRIPTION: Greg Knieriemen (@knieriemen, Director of Technology Evangelism at @NetApp) talks about the growing importance of data to reshape business, emerging data trends and standards, working around 5G networks and more.SHOW SPONSOR LINKS:UpCloud - World’s fastest cloud serversUpCloud (promo) - Sign up for free, receive a $50 credit and try us out!MongoDB Homepage - The most popular database for modern applicationsMongoDB Atlas - MongoDB-as-a-Service on AWS, Azure and GCPDatadog Homepage - Modern Monitoring and AnalyticsTry Datadog yourself by starting a free, 14-day trial today. Listeners of this podcast will also receive a free Datadog T-shirt[DONUT RUN DONATIONS] SHOW NOTES:Go Your Way (podcast)Enterprise Te.ch (Greg’s newsletter)California’s new Data Privacy Law - California Consumer Privacy Act (CCPA)Topic 1 - Welcome back to the show. For anyone that doesn’t know you, tell us a little bit about your background, and some of the areas you focus on now.Topic 2 - Let’s start by having an executive-level discussion about “data”. Everyone loves talking about developers and applications, but how do you explain to an executive about the value of managing data as a core business asset? What makes data so valuable in 2020?Topic 3 -  The world used to be “big servers” + “big relational database” + “storage array”. Now that model is blow up into 100s of distributed elements, lots of technology choices, lots of locations. Is there a way to think about data in these distributed models? Topic 4 - What are some of the emerging data standards that people should be aware of, or learning more about? [alternatives to 5G]Topic 5 - Data moved from a cost “liability” a while ago, to an “asset” as we started talking about Data Lakes and AI/ML models. Have you found any metrics or conversations that people understand (in the “analytics context”) about how we justify keeping data - or trying to acquire more data. Topic 6 - We’ve got cheaper ways to store data, faster networks coming (e.g. 5G, alternative options) and cheaper/faster CPUs (e.g. ARM). When these dynamics change, new things emerge that we’ve never thought of before. Are you starting to hear people discuss interesting new business ideas yet? FEEDBACK?Email: show at thecloudcast dot netTwitter: @thecloudcastnet

SHOW: 430DESCRIPTION: Aaron and Brian discuss the biggest trends of 2019, and make bold cloud computing predictions for 2020. SHOW SPONSOR LINKS:Datadog Homepage - Modern Monitoring and AnalyticsTry Datadog yourself by starting a free, 14-day trial today. Listeners of this podcast will also receive a free Datadog T-shirtMongoDB Homepage - The most popular database for modern applicationsMongoDB Atlas - MongoDB-as-a-Service on AWS, Azure and GCP[DONUT RUN DONATIONS] [FREE] Try an IT Pro ChallengePODCAST BUSINESS: Why are Aaron and Brian both on the show recently?Krispy Kreme Challenge (“The Donut Run”) fundraisingAnnouncing “The Cloudcast Basics” - coming in early 2020!! CLOUD NEWS OF THE WEEK: The Cloudcast in 2019:Over 1.6M+ listens, up 40% YoYGuest Acquisitions: (total: 9) Cloudability (Apptio), Bromium (HPE), Docker (Mirantis), Shippable (JFrog), SignalFX (Splunk), Pivotal (VMware), NGINX (F5), ParkMyCloud (Turbonomics), Twistlock (Palo Alto) - also had 9 in 2018IPO: DatadogTRENDS and MAJOR STORIES from 2019:Public Cloud CAPEX comparisons (through 2018)Data of AWS trends in Revenues, Growth Rates, Operating Margins, Operating IncomeAWS continues to lead in revenues, but their revenue growth has been slowing (as a YoY and QoQ%) for the last 5-6 months. Increased competition from Azure.Azure won the US DoD JEDI contract. Google continues to be 3rd or 4th cloud, with Alibaba Cloud often ranked 3rd.IBM closed the $34B acquisition of Red HatVMware made 8 acquisitions in 2019 - Bitnami, Pivotal. AVI Networks, Carbon BlackLots of discussion about large public cloud spending by web scale companies (Salesforce, Apple, Spotify, etc.)Several of the Gig-Economy companies continued to struggle in finding a profitable business model - Uber, Lyft, WeWork, DoorDash, etc.2020 PREDICTIONSBrian:We’ll start talking about GitHub as one of the major cloud platforms, in the same way we do AWS, Azure and Google. We’re going to start seeing more and more vertical-centric AI/ML companies emerge, that curate data and provide insights-as-a-service. We’re going to start seeing companies that offer distributed versions of the large “monolithic” systems of today (Core Banking, ERP, etc.) that lets new companies and business models emerge.Aaron:The “trough of disillusionment” will hit Kubernetes, and it will be fine…Serverless will get a new name and will hit strideGitLab and HashiCorp will have breakout yearsSTARS WARS DISCUSSION:For the 1st time ever, we indulge Aaron and talk a little bit of sci-fi on the

SHOW: 429DESCRIPTION: Aaron and Brian discuss the biggest trends of the 2010s decade, and answer reader mailbag questions from 2019. SHOW SPONSOR LINKS:MongoDB Homepage - The most popular database for modern applicationsMongoDB Atlas - MongoDB-as-a-Service on AWS, Azure and GCPDatadog Homepage - Modern Monitoring and AnalyticsTry Datadog yourself by starting a free, 14-day trial today. Listeners of this podcast will also receive a free Datadog T-shirt[DONUT RUN DONATIONS] [FREE] Try an IT Pro ChallengeTRENDS FROM THE DECADE (2010-2019):Hype or Not?: OpenStack, HCI, Docker, Cloud Foundry (or any PaaS), Kubernetes, Blockchain, AI/ML, ServerlessThe re-invention of Microsoft (Azure, GitHub, Open Source, dropping failed efforts (Nokia, Windows Mobile)It’s hard to believe that Google is still only 3rd or 4th in public cloud (and it’s leadership is questioning its future)The rise of open-source in the Enterprise, and all the challenges this created across the industryThe rise of developer-lead decision-making (shadow IT, mobile apps, decentralized decision-making, serverless as NoOps)Everyone still sucks at security, but now it’s a headline problem10 years of DevOps discussions and most companies still haven’t figured out that it’s org-charts and culture.The shift from Config-Mgmt (Ansible, Chef, Puppet, Salt) to things like Immutable infrastructure, Declarative schedulers, GitOps, etc.)READER MAILBAG QUESTIONS:Question 1 - If you had to advise someone that’s 25, 35 and 45 about “career path” in IT, what tips might you give them? - Thomas T.Question 2 - You guys see a lot of different technologies through the interviews. Which new ones do you think have longevity and which ones are potentially overhyped? - Michael C.Question 3 - What “disruptive technology” is a more important skill to learn? - Erika B.Question 4 - What the most misunderstood concept in IT today? Seema J.Question 5 - Why do you think our industry is so obsessed with the idea that the new technology will completely eliminate the old technology? Has that ever really happened? - Jacob H.