This is where I publish writing on infrastructure, automation, digital systems, and the practical lessons that come from building and maintaining them.
This is a compact troubleshooting post: if WSL complains about an unknown wsl2.autostart key, the fix is not another config flag. The practical answer is to use Task Scheduler or the Startup folder instead.
This draft rewrites a real mail operations note into a cleaner guide for improving send-only SMTP deliverability by wiring Postfix to OpenDKIM, generating per-domain keys, and publishing the required DNS records.
This draft keeps the operations side of a monitoring rollout intact: install kube-prometheus, expose the main services, add alerting configuration, and adjust the RBAC when the cluster needs broader access than the defaults provide.
This draft captures the practical flow for building a bare-metal Kubernetes cluster with MAAS and DeepOps: commission the nodes, deploy the operating system, prepare the inventory, tune the Kubernetes variables for the environment, and run the cluster build with enough observability to make the result usable.
This draft turns a raw build note into a practical cluster setup guide: wipe the storage disks carefully, prepare the network, deploy Kubernetes with DeepOps, add MetalLB, and leave the cluster in a state where Ceph and monitoring can be layered on cleanly.
This draft captures a practical bare-metal monitoring setup: install kube-prometheus, remove or adjust the network policies that block outside access, add MetalLB, convert the services to LoadBalancer, and patch the RBAC where the stock install is not enough for the cluster.
This draft captures the architecture and operational flow of a FusionAuth cluster build: prepare the Kubernetes base, add MetalLB, deploy PostgreSQL, choose and validate the search backend, and work through the application-side dependency checks before trying to restore or migrate real data.
This draft keeps the useful part of a migration note intact: back up the old Uptime Kuma data, install the chart, restore the data into the new deployment, switch the service to LoadBalancer, and verify the endpoint from outside the cluster.
This draft keeps the useful migration flow from a sensitive work log: prepare storage, install the package, preserve the SSH host identity, move the backups and secrets, restore GitLab cleanly, and validate the instance before you start trusting it again.
This draft turns a storage incident note into a troubleshooting pattern: reproduce the issue with a minimal PVC and pod, inspect the events and Longhorn volume state, and use that evidence to separate application problems from storage attachment problems.