Software Daily

Software Daily

By SoftwareDaily.comNewsTech News
Download on the App Store

Software Daily episodes

  • Infrastructure Wars with Sheng Liang

    Sheng Liang was the lead developer on the original Java Virtual Machine. Today he works as the CEO of Rancher Labs, a company building a platform on top of Kubernetes. Sheng joins the show to discuss his experiences in the technology industry.

    The container orchestration wars had many victims. The competing standards for how an enterprise should manage its numerous containers caused several companies to go down a path where they were building infrastructure which eventually had to be replaced.

    As Sheng discusses in today’s episode, the container orchestration wars almost killed his company. Rancher was originally built on top of a different container orchestrator, and the migration to Kubernetes required a massive rewrite of the Rancher platform.

    The container orchestration wars were not the first technology battle that Sheng has seen in his career–and it won’t be his last. In today’s show, we discuss the nature of technology wars. Are they necessary? How can a software company minimize the damage caused by a war between competing standards?

    Sheng was an excellent guest and we didn’t cover nearly as many subjects as I wanted to, so we will have to do another show in the future!

    0 min
  • Kubernetes Operators with Rob Szumski

    Kubernetes has made distributed systems easier to deploy and manage. As Kubernetes has become reliable, engineers have started to look for higher level abstractions we can define on top of Kubernetes.

    An operator is a method of packaging, deploying, and managing a Kubernetes application.

    Operators are useful for spinning up distributed systems such as Kafka, Redis, or MongoDB. These data systems are complicated, stateful applications with lots of failure domains. The operator framework enables a developer to deploy one of these complicated applications with less fear of the system crashing, or entering an erroneous state.

    Rob Szumski is an engineer at Red Hat. He joins the show to discuss Kubernetes, the operator pattern, and his time at CoreOS, which was acquired by Red Hat.

    0 min
  • Kubernetes Operators with Rob Szumski

    Kubernetes has made distributed systems easier to deploy and manage. As Kubernetes has become reliable, engineers have started to look for higher level abstractions we can define on top of Kubernetes.

    An operator is a method of packaging, deploying, and managing a Kubernetes application.

    Operators are useful for spinning up distributed systems such as Kafka, Redis, or MongoDB. These data systems are complicated, stateful applications with lots of failure domains. The operator framework enables a developer to deploy one of these complicated applications with less fear of the system crashing, or entering an erroneous state.

    Rob Szumski is an engineer at Red Hat. He joins the show to discuss Kubernetes, the operator pattern, and his time at CoreOS, which was acquired by Red Hat.

    0 min
  • Render: High Level Cloud with Anurag Goel

    Cloud computing was popularized in 2006 with the launch of Amazon Web Services. AWS allowed developers to use remote server infrastructure with a simple set of APIs. But even with AWS, it was still not simple to deploy and manage a web application.

    In 2007, Heroku launched a platform built on top of AWS. Heroku focused on the developer experience by optimizing for users who were deploying Ruby on Rails applications. Since then, Heroku has expanded into other forms of managed infrastructure, including other application frameworks like NodeJS, and databases like Postgres.

    Heroku was the first popular “layer 2” cloud provider. Twelve years later, it is probably still the most popular. But there have been many other cloud providers built on top of AWS, including Netlify, Zeit, Spotinst, and Firebase.

    Layer 1 cloud providers are Google Cloud, AWS, Azure, Digital Ocean, and other raw infrastructure providers. These companies provide a great service in their low cost, commodity infrastructure. But the layer 1 providers are not optimizing for developer experience. They need to cater to a broad set of developers, some of whom want to work at a low level.

    A layer 2 cloud provider can build an opinionated solution that serves a subset of the overall cloud market particularly well.

    Render is a layer 2 cloud provider that optimizes for specific developer workflows, such as deploying a NodeJS web server, a static site, or a Docker container. Anurag Goel is the founder of Render, and he joins the show to discuss the strategy and the economics of Render. Anurag was also one of the early employees at Stripe, and he discusses his experience and learnings from working at the company.

    0 min
  • Elegant Puzzle with Will Larson

    Software engineering is an art and a science. To manage engineers is to manage artists and scientists.

    Software companies build practical tools like payment systems, messaging products, and search engines. Software tools are the underpinnings of our modern lives. You might expect this core infrastructure which modern humans rely on to have been constructed with pure formulaic rigor.

    But the best software tools are not built within a totally defined process. Software is built through messy iteration. When a piece of software looks pristine, that is often a function of how many mistakes have been made, and then subsequently corrected for.

    There is no fixed process for how to build good software.

    As our tools get better, we have to update our software engineering practices to utilize those new tools. We have to rethink the style that we are working in. We have to discard old tools and procedures in order to pick up the new ones, and have higher leverage.

    As an organization scales, the structure of the organization needs to be modified. Team members need to be reallocated. Checks and balances need to be put in place. Rules and cultural practices need to codified, because a larger organization cannot have ties broken by an individual.

    Software is built by humans, and every management decision must be considered in the light of human psychology. When we change a line of code, the code does not get emotional about being altered. But the same cannot be said of humans. Even a minor conversation between an engineering manager and a direct report can have lasting implications.

    Will Larson is the author of An Elegant Puzzle: Systems of Engineering Management. He works on Foundation Engineering at Stripe, and has worked in engineering management at Uber, Digg, and other software companies. Elegant Puzzle provides strategies, tactics, and ruminations on software development. Will joins the show to explore the multifaceted subject of engineering management.

    0 min
  • Elegant Puzzle with Will Larson

    Software engineering is an art and a science. To manage engineers is to manage artists and scientists.

    Software companies build practical tools like payment systems, messaging products, and search engines. Software tools are the underpinnings of our modern lives. You might expect this core infrastructure which modern humans rely on to have been constructed with pure formulaic rigor.

    But the best software tools are not built within a totally defined process. Software is built through messy iteration. When a piece of software looks pristine, that is often a function of how many mistakes have been made, and then subsequently corrected for.

    There is no fixed process for how to build good software.

    As our tools get better, we have to update our software engineering practices to utilize those new tools. We have to rethink the style that we are working in. We have to discard old tools and procedures in order to pick up the new ones, and have higher leverage.

    As an organization scales, the structure of the organization needs to be modified. Team members need to be reallocated. Checks and balances need to be put in place. Rules and cultural practices need to codified, because a larger organization cannot have ties broken by an individual.

    Software is built by humans, and every management decision must be considered in the light of human psychology. When we change a line of code, the code does not get emotional about being altered. But the same cannot be said of humans. Even a minor conversation between an engineering manager and a direct report can have lasting implications.

    Will Larson is the author of An Elegant Puzzle: Systems of Engineering Management. He works on Foundation Engineering at Stripe, and has worked in engineering management at Uber, Digg, and other software companies. Elegant Puzzle provides strategies, tactics, and ruminations on software development. Will joins the show to explore the multifaceted subject of engineering management.

    0 min
  • Stripe Machine Learning Infrastructure with Rob Story and Kelley Rivoire

    Machine learning allows software to improve as that software consumes more data.

    Machine learning is a tool that every software engineer wants to be able to use. Because machine learning is so broadly applicable, software companies want to make the tools more accessible to the developers across the organization.

    There are many steps that an engineer must go through to use machine learning, and each additional step inhibits the chances that the engineer will actually get their model into production.

    An engineer who wants to build machine learning into their application needs access to data sets. They need to join those data sets, and load them into a machine (or multiple machines) where their model can be trained. Once the model is trained, the model needs to test on additional data to ensure quality. If the initial model quality is insufficient, the engineer might need to tweak the training parameters.

    Once a model is accurate enough, the engineer needs to deploy that model. After deployment, the model might need to be updated with new data later on. If the model is processing sensitive or financially relevant data, a provenance process might be necessary to allow for an audit trail of decisions that have been made by the model.

    Rob Story and Kelley Rivoire are engineers working on machine learning infrastructure at Stripe. After recognizing the difficulties that engineers faced in creating and deploying machine learning models, Stripe engineers built out Railyard, an API for machine learning workloads within the company.

    Rob and Kelley join the show to discuss data engineering and machine learning at Stripe, and their work on Railyard.

    0 min
  • Stripe Machine Learning Infrastructure with Rob Story and Kelley Rivoire

    Machine learning allows software to improve as that software consumes more data.

    Machine learning is a tool that every software engineer wants to be able to use. Because machine learning is so broadly applicable, software companies want to make the tools more accessible to the developers across the organization.

    There are many steps that an engineer must go through to use machine learning, and each additional step inhibits the chances that the engineer will actually get their model into production.

    An engineer who wants to build machine learning into their application needs access to data sets. They need to join those data sets, and load them into a machine (or multiple machines) where their model can be trained. Once the model is trained, the model needs to test on additional data to ensure quality. If the initial model quality is insufficient, the engineer might need to tweak the training parameters.

    Once a model is accurate enough, the engineer needs to deploy that model. After deployment, the model might need to be updated with new data later on. If the model is processing sensitive or financially relevant data, a provenance process might be necessary to allow for an audit trail of decisions that have been made by the model.

    Rob Story and Kelley Rivoire are engineers working on machine learning infrastructure at Stripe. After recognizing the difficulties that engineers faced in creating and deploying machine learning models, Stripe engineers built out Railyard, an API for machine learning workloads within the company.

    Rob and Kelley join the show to discuss data engineering and machine learning at Stripe, and their work on Railyard.

    0 min
  • Software Businesses with Hiten Shah

    The software market changes every year.

    As individuals and enterprises become more willing to buy software, there are new markets for entrepreneurs to sell software to. Good software has high margins and high retention, so even a niche software business can prove profitable.

    As software spreads across the world, developing countries are showing a willingness to buy the same software tools as the developed countries, making the niche software businesses even more successful than expected.

    It’s a great time to start a software business. But the fundamental challenges faced by an entrepreneur have not gone away. An entrepreneur must find the necessary capital, build an initial product, iterate on that product, market to their audience, sell to their customers, and retain their users.

    Hiten Shah is an entrepreneur and investor who has started several businesses, both successful and unsuccessful. He is also the host of the Startup Chat podcast, a show about strategies and tactics for software businesses. Hiten joins the show to discuss his experience in the software industry, including lessons on financing, product development, and the future of software products.

    0 min
  • Kubernetes Vision with Joe Beda

    Google Cloud was started with a vision of providing Google infrastructure to the masses.

    In 2008, it was not obvious that Google should become a cloud provider. Amazon Web Services was finding success among startups who needed on-demand infrastructure, but the traditional enterprise market was not yet ready to buy cloud resources.

    Googlers liked the idea of becoming a cloud provider. But was it the right time to enter the market? Google’s advertising business was a large and growing cash cow. Executives within Google were not sure how much capital and effort should be allocated into an infrastructure business.

    When Google decided to go into the cloud business, Joe Beda was one of the engineers who helped lead the effort, and joins the show as today’s guest.

    Google’s internal server infrastructure is managed by Borg, a system for allocating resources to applications. Google Cloud runs on Borg, and there were a number of early engineering challenges to building the necessary functionality into Borg for running a cloud provider on top of it.

    One example of a technical challenge that Google faced was the refactoring of Borg to run Google Cloud workloads.

    The requirements for public infrastructure are different than those of internal Googlers. Inside of Google, developers deploy their applications to containers running on bare metal. Outside of Google, developers want to create virtual machines. Borg needed to be refactored in order to instantiate VMs.

    Google solved this technical problem, as well as many other challenges, and Google Cloud slowly gained momentum in the market. But AWS remained the default choice for profitable enterprise workloads. It wasn’t until the container orchestration wars that Google found an opportunity to jump on a market segment that offered strong differentiation.

    By open sourcing Kubernetes and presenting a clear vision for where the project was going, Google shifted the battlefield of the public cloud toward a competitive landscape where it has many advantages. Kubernetes also provided many other technology companies with an opportunity to get into the cloud market, creating a collaborative, multi-company ecosystem that has accelerated the pace of software faster than anyone expected.

    Joe Beda has been instrumental in the evolution of the cloud native ecosystem. In today’s episode, Joe gives his memories on Google Cloud, Kubernetes, and his Kubernetes company Heptio, which he sold to VMware.

    0 min

About Software Daily

From the publisher's feed

Technical interviews about software topics.