Blog
Practical notes from our cloud and DevOps work: upgrades, releases, security, compliance and cloud cost.
- FinOps
How to reduce AWS costs without affecting production
A safe order for cutting an AWS bill: read-only findings first, then non-production, storage, compute and commitments last, with a rollback for each step.
· 8 min readRead → - Security
47-day TLS certificates: what changes for teams hosting many sites
TLS certificate lifetimes drop to 200 days in 2026 and 47 days by 2029. What that means if you host dozens or hundreds of sites, and how to automate renewals.
· 6 min readRead → - Databases
Upgrading Amazon RDS PostgreSQL 14 to 17: a practical runbook
Plan an RDS PostgreSQL 14 to 17 upgrade: snapshot rehearsals, Blue/Green deployments, the SSL default that breaks apps, extensions and post-upgrade checks.
· 7 min readRead → - CI/CD
Keyless CI/CD: deploy to AWS from GitLab or GitHub with OIDC
Stop storing AWS keys in CI. How OIDC lets GitLab CI and GitHub Actions assume an IAM role with short-lived credentials, and how to lock it down.
· 6 min readRead → - Containers
Scaling an ECS Fargate worker on SQS queue depth, down to zero
How to autoscale an ECS Fargate queue worker on SQS backlog per task, scale to zero when idle, and handle shutdowns, retries and dead-letter queues safely.
· 6 min readRead → - FinOps
Why right-sizing AWS does not always cut your bill straight away
Savings Plans and Reserved Instances bill every hour, used or not. How to time right-sizing with commitment expiry so the savings actually show up.
· 5 min readRead → - Compliance
SOC 2 for a small SaaS team: the evidence auditors actually ask for
A plain-English guide to SOC 2 Type I and Type II for small SaaS teams: trust criteria, the evidence auditors request, and the AWS controls that produce it.
· 7 min readRead →
