The Future of Cloud Engineering Evolves

The 48th IT Press Tour had the opportunity to meet with Joe Duffy, founder, and CEO of the Pulumi cloud engineering platform. Their mission is to democratize the cloud as the modern cloud changes everything, and yet little has changed when it comes to working with the cloud. Joe and his co-founder, Eric Rudder, were both at Microsoft. Joe ran Microsoft's Developer Tools and Platform strategy while Eric was an EVP and Chief Technical Strategy Officer.

Challenges of the Modern Cloud

Developers treat the cloud as an afterthought, while infrastructure teams struggle with solutions that do not scale. According to a recent CNCF survey, 50% of developers wait up to one month to get infrastructure access, and security continues to be an afterthought resulting in insecure code and applications.

Yesterday's tools and practices don't scale in the modern cloud era. The scale comes in many forms. Architecture goes from tens of services and environments to thousands. Code bases from tens of IaC LOCs to hundreds of thousands. Code has gone from shipping quarterly to daily. Agile teams scale up and down as needed and now involve dev plus ops rather than just ops. Security needs to be integrated during construction.

Today, all software is a cloud software, all developers are cloud developers, and forward-thinking people realize infrastructure enables innovation. Joe sees several inflection points in cloud adoption: 

1. When AWS launched S3; 
2. Containers as a way to package code in a cloud-native way; 
3. Kubernetes; and 
4. Cloud services with Lambda. 

Businesses are reshaping themselves with software thanks to the cloud. Devs are getting their hands dirty and thinking about the cloud because it allows them to deliver more value more quickly. Large enterprises are putting together platform teams to empower devs without the need to be cloud experts.

The Solution

Pulumi tapped into decades of great languages to stand on the shoulders of giants. They have taken everything devs and engineers love about infrastructure as code (IaC) and married that with everything they love about languages. They have a point along the spectrum for every engineer. Their favorite IDE just works—  all their tools work too. 

One platform for multi-cloud. Workflows are standard across any cloud—  public, private, or hybrid. Any persona—  developer, infrastructure expert, security engineer. Standardized while embracing the uniqueness of each cloud. 

The platform integrates and extends existing ecosystems with hundreds of partner integrations, packages, and dozens of CI/CD systems to unify application and infrastructure delivery. They support any modern cloud workloads—  serverless, containers, and VMs. It facilitates the build, deploy, manage, and secure development process using a consistent approach and eliminates the custom integration of dozens of tools. 

Shift Left With Guardrails

A central cloud engineering platform defines consistent workload, architecture, and best practices. This enables developers (Application, Fulls Stack, Systems Engineers) to build cloud-native apps, provision self-service cloud applications and infrastructure, and leverage the cloud to deliver new value-added capabilities and services. 

The platform enables infrastructure experts (IT, Ops, DevOps, SREs, Systems Engineers, and Production Engineers) to provision, automate, and manage infrastructure and operations. Security teams (Security Engineer, Security Architect, Compliance Experts, DevSecOps) to secure the entire stack and ensure governance. 

The cloud engineering platform is open source and provides individual building blocks for all cloud and resources as well as complete, modern, multi-cloud best practices architectures. In addition, there are cloud-specific libraries for common patterns and practices. This enables users to go from manual steps and markup to built-in best practices. 

Architecture templates are baseline blueprints for the most common cloud architectures. Users can start with the most common infrastructure as code projects with built-in best practices. On any cloud; AWS, Azure, GCP, Kubernetes, et al. Containerized microservices, serverless applications, static websites, VMs, base networking layers, and managed Kubernetes clusters. In any language; TypeScript, Python, Go, C#, Java, YAML, etc.

Features include:

• Total visibility and control.
• Federated identity and group management; SSO with SAML and automatic group synchronization.
• Multi-cloud policy as code to enforce security, compliance, and cost controls
• Built-in, encrypted secrets management.
• Automation API to build internal platforms or to offer cloud service provisioning.

Organizations using Automation API manage more than 10X the cloud infrastructure resources per engineer versus traditional CLIs. Cockroach Labs built a SaaS product for managed databases that enable their engineers to handle 200X resources compared to the traditional IaC user. 

Customer Feedback and Use Cases

Atlassian reduced developers' time spent on maintenance by 50%. "Our team quickly and easily built a repeatable pipeline that deploys development environments for more than 100 team members worldwide. In addition, the benefits of working in languages we know can't be overstated. That just made moving to the Pulumi Cloud Engineering Platform an obvious choice for the team."

Virtual Gaming Worlds is an enterprise customer. "I am blown away by the power of Pulumi. It is truly amazing, and I love how it empowers all engineers to make infra part of their inner loop. Being able to work in a programming language that you are familiar with and not have to learn something completely different really helps. It's also a really fun way of getting familiar with the cloud provider services."

Mercedes Benz used Pulumi to empower its developers to move faster and scale hundreds of Kubernetes clusters worldwide. "Pulumi solves the multi-cloud problem in exactly the way I was looking for." 

Use Cases:

• Snowflake —  Simple to manage Kubernetes on AWS, Azure, and Google.
• Hyland —  98% reduction in lines of infrastructure code.
• Atlassian —  50% reduction in time spent on cloud maintenance.
• Panther —  10X faster deployments and more frequent testing.
• Fauna —  Reduced time to market from weeks to days.
• Green Park —  100% of developers use IaC leading to faster velocity.
• Mercedes Benz —  Developers can self-service approved infrastructure.
• Lemonade —  Automatically enforce policies for security and compliance.

Key Takeaways

• It was built for the entire team —  Developers and infrastructure experts alike.
• Standard multi-cloud workflow —  Public, private, or hybrid.
• Developer adoption and productivity — Use favorite languages and tools.
• Sharing and reuse —  Capture best practices in reusable assets.
• Embrace cloud-native architectures —  Containers, serverless, managed services, and VMs.
• Scale with complexity and business needs —  From tens, hundreds, and thousands of resources and environments.
• Ship faster with confidence —  In hours versus weeks.