AWS Organizations SCPs: The Guardrail That Beats a Strongly Worded Wiki
Service control policies in a multi-account org: deny lists, allowed regions, and why SCPs do not replace IAM identity policies.
Category
20 articles · 67 min of reading
Service control policies in a multi-account org: deny lists, allowed regions, and why SCPs do not replace IAM identity policies.
AWS Graviton, multi-arch containers, AArch64 JDK builds, and the native libraries that fail first when you leave x86.
CloudFront, Cloud CDN, and Azure Front Door style signatures: expiry, prefix scope, and the leak that is a URL in a log file.
Zonal VMs, zone-redundant services, and regional pairs: what a datacenter fire takes, and what a region-level event still takes.
Singh et al. SIGCOMM: hoverboards, flow programming, and why GCP's VPC can be software without looking like a software switch on your VM.
Enclaves, vsock, attestation documents, and KMS Decrypt that only the enclave can call — plus the operational limits that keep this from being 'just a VM'.
Which of Heroku's 2011 factors still prevent outages, which ones aged (local disk, log sockets), and how to apply them without cargo-culting config vars.
How a service with a broad role gets tricked into acting on the wrong resource, and the sts:ExternalId / source ARN patterns that stop it.
What read-after-write means on S3 today, why list-after-delete still bites, and how to design keys, retries, and multipart uploads around that.
Why naive CPU-based autoscaling fails under real traffic patterns, and the metric choices and tuning that make autoscaling respond correctly instead of thrashing.
A practical walkthrough of VPC networking concepts, from subnets and route tables to security groups, aimed at developers who deploy but don't design networks.
EKS, GKE, and AKS all promise to take Kubernetes operations off your plate, but the actual division of responsibility is narrower than most teams assume.
Edge functions promise lower latency by running closer to users, but they come with real constraints on runtime and state. Here's where they genuinely help.
Multi-region architecture is sold as pure resilience, but the real trade-offs around data consistency, latency, and operational cost rarely make it into the pitch.
Why Terraform state is the most fragile part of most infrastructure setups, and the locking, backend, and workspace patterns that keep it from becoming a liability.
The real architectural differences between message queues and event streams, and why picking the wrong one causes problems that only show up under load.
Concrete cloud cost optimization tactics, from rightsizing to commitment discounts, ranked by effort versus savings so you know where to start.
Least privilege is easy to state as a principle and hard to implement well. Here's how to actually scope IAM policies without breaking every deploy.
Key design patterns for using S3-style object storage well, from key naming schemes to lifecycle policies, and the mistakes that cause throttling at scale.
What causes cold starts in serverless functions, why runtime and package size matter more than most tuning knobs, and the mitigations that actually move the needle.