Terraform migrate 1.1 adds VCS workspace support and enhanced GitOps (2 minute read)
Terraform migrate 1.1 now supports VCS workspaces and expanded Git capabilities, enabling automated runs via version control systems and GitOps practices. The update allows users to configure Terraform migrate to work from existing repository branches, with features like Git token validation, Git operations, and pull-request creation becoming more adaptive.
|
|
Migrating Uber's Compute Platform to Kubernetes: A Technical Journey (10 minute read)
Uber migrated all shared stateless workloads from Mesos to Kubernetes to leverage industry-standard tooling, improve scalability, and ensure infrastructure resilience. The transition was completed by mid-2024. It required overcoming large-scale operational challenges, rebuilding ecosystem integrations, achieving feature parity, and implementing custom solutions like artifact preservation, gradual scaling, and faster rollback mechanisms.
|
Building Industrial Strength Software without Unit Tests (8 minute read)
Unit tests can be cumbersome and unnecessary for certain projects. Developers should instead use transcript tests, which are deterministic, markdown-based integration tests that interact with a system's external interface and can double as documentation and bug reports. Unison successfully used this approach by scripting full workflows and capturing outputs, proving it to be a maintainable and developer-friendly alternative to traditional testing practices.
|
|
KAI Scheduler (GitHub Repo)
KAI Scheduler, a Kubernetes native scheduler, optimizes GPU resource allocation for large-scale AI and machine learning workloads by dynamically allocating resources and ensuring fairness. Designed to manage extensive GPU clusters, KAI Scheduler supports the entire AI lifecycle and can be installed alongside other schedulers.
|
Suna (GitHub Repo)
Suna, a fully open-source AI assistant, was released to help users automate tasks through natural conversation, browser automation, file management, and API integrations. The AI agent consists of a Python/FastAPI service and Next.js/React application. It supports self-hosting.
|
|
Cloud Realities Are Slowing AI Ambitions (4 minute read)
AI workloads, especially with the rise of Agentic AI, are straining traditional cloud configurations, requiring enterprises to re-architect systems for greater interactivity, lower latency, and resilience. To maintain performance, businesses must reassess cloud infrastructure, traffic routing, and service dependencies to meet the demands of AI at scale.
|
Tests should not contain logic (2 minute read)
Tests should avoid logic to prevent masking bugs in the code under test—duplicating the same flawed logic in a test will lead to false positives. Use parametrized tests instead to concisely cover multiple cases with explicit inputs and expected outputs, ensuring clarity and correctness without control flow in the test itself.
|
|
|
Love TLDR? Tell your friends and get rewards!
|
|
Share your referral link below with friends to get free TLDR swag!
|
|
|
|
| Track your referrals here. |
|
Want to advertise in TLDR? 📰
If your company is interested in reaching an audience of devops professionals and decision makers, you may want to advertise with us.
Want to work at TLDR? 💼
Apply here or send a friend's resume to jobs@tldr.tech and get $1k if we hire them!
If you have any comments or feedback, just respond to this email!
Thanks for reading,
Kunal Desai & Martin Hauskrecht
|
|
|
|
