
How to Use Prometheus and node exporter for Server Metrics
Install node exporter as a systemd service, scrape it with Prometheus, write PromQL for CPU, memory and disk, and keep the port private.
Tag
11 posts

Install node exporter as a systemd service, scrape it with Prometheus, write PromQL for CPU, memory and disk, and keep the port private.

Distribute a script with GitHub Releases: tags, notes, attached files and checksums, the gh CLI, an installer that verifies and update checks.

Automate releases and changelogs in CI from a git tag: trigger on tags, generate notes with gh or git-cliff, attach artifacts and keep it reproducible.

Run capacity checks from the shell: disk and inode use, memory available, load against core count, a growth estimate and a simple CSV log for trends.

Write a runbook for common server problems: a fixed template with symptom, check, fix and verify, copy-paste commands, an owner and a way to test it.

Use git tag for releases: lightweight versus annotated tags, signing, pushing to a remote, listing and deleting, and naming versions with semantic versioning.

Deploy from CI over SSH with a dedicated key limited by authorized_keys options, a pinned host key, a single deploy command and no StrictHostKeyChecking off.

Use Helm from the command line: add repos, install and upgrade releases, values files, template and diff before applying, history, rollback and uninstall.

Use Makefile deploy targets to wrap build, test, release and rollback commands, with phony targets, variables, a production guard and a self-documenting help.

What is GitOps: infrastructure described in Git and applied by an agent. See the pull model, the tools, how a small team can start, and when it is overkill.

Use simple feature flags with environment variables, a config file and a small script, roll out gradually, remove old flags, and know when to use a service.