Kubernetes overview
The project’s own explanation of its purpose, capabilities, historical context, and boundaries.
Selected reads
One line through
the useful material.
A direct reading list for understanding the systems behind the work. No courses, stages, or intermediate pages. Every title opens the source.
Kubernetes
How does Kubernetes work?
Kubernetes components
Maps the API server, etcd, scheduler, controller manager, kubelet, kube-proxy, and container runtime onto a cluster.
What happens when Kubernetes edition
A source-linked execution trace from a kubectl command to a running Pod, including the control plane and node-side work.
vLLM
How does vLLM serve a language model?
Inside vLLM: anatomy of a high-throughput LLM inference system
The strongest end-to-end explanation of request processing, continuous batching, PagedAttention, prefix caching, scheduling, decoding, and throughput trade-offs.
vLLM architecture overview
Maps the API server, engine core, scheduler, cache manager, and GPU workers onto the current process architecture and execution path.
Efficient memory management for large language model serving with PagedAttention
Introduces the memory problem, PagedAttention, cache sharing, scheduling consequences, and the evaluation behind the original vLLM system.
Permission engines
How should a permission engine model authorization?
Guide to Attribute Based Access Control definition and considerations
Defines ABAC’s terms, policy model, enterprise considerations, and relationship to role-based access control without tying the model to a vendor.
Zanzibar: Google’s consistent, global authorization system
The foundational system paper behind modern ReBAC engines, covering the relation model, tuple storage, consistency, distributed evaluation, and production scale.
Authorization API 1.0
Defines a model-neutral API between applications and authorization systems, including evaluation, batch evaluation, search, context, and decision responses.
Argo CD
How does Argo CD reconcile Git with Kubernetes?
Argo CD architectural overview
The shortest authoritative map of Argo CD’s components and the application controller that compares live state with the target state in Git.
Sync phases and waves
Connects PreSync, Sync, PostSync, and SyncFail hooks with wave ordering and the health-based progression of a deployment.
Apache Iceberg
How does an Apache Iceberg table work?
Apache Iceberg table specification
The canonical description of the metadata graph, sequence numbers, snapshots, optimistic commits, partition specs, schemas, and row-level deletes.
Evolution in Apache Iceberg
Concentrates the schema, partition, sort-order, and property changes that Iceberg applies as metadata operations.
Teleport
How does Teleport replace infrastructure credentials?
Teleport architecture
The authoritative component and trust model, including how clients authenticate and how the proxy reaches registered infrastructure without distributing static credentials.
Access requests
Covers the time-bound elevation path that turns standing administrative access into requested, reviewed, expiring access.
Keyless CI/CD
How does OIDC remove cloud credentials from CI/CD?
OpenID Connect Core 1.0
The protocol source for ID tokens, claims, validation, discovery assumptions, and the distinction between authentication and OAuth authorization.
OpenID Connect in GitHub Actions
Shows the workflow side of the exchange, including token claims, cloud trust conditions, short-lived credentials, and the removal of stored deployment secrets.
Create an OIDC identity provider in IAM
Covers the AWS trust object, issuer metadata, certificate chain, audience configuration, and the relationship between the provider and role trust policy.