Session Overview: The Compute Spectrum
The second session in GDG on Campus PUP’s Cloud Solutions Series picked up where onboarding left off — framed around a central architectural decision: When you have an application idea, how do you systematically choose where to run it?
The workshop structured cloud compute along the control spectrum:
- Full Infrastructure Control (IaaS) — Virtual machines with customized OS kernels, storage attachments, and networking rules.
- Simplicity & Speed (PaaS) — Managed runtime platforms that abstract operating system maintenance.
- Event-Driven Flexibility (FaaS & Serverless Containers) — Ephemeral micro-billing runtimes scaling down to zero.
Hands-On Demonstrations: Compute Engine & Cloud Run
I covered both ends of the control continuum:
- Compute Engine Deep Dive — Explored what a virtual machine actually represents under hypervisors, machine series trade-offs, persistent disks, firewall rules, and cost optimization levers. Followed immediately with a live console walkthrough configuring NGINX on Debian.
- Cloud Run & Containerized Deployments — Contrasted container packaging against raw VMs, explained the 2025 request-based billing model, and demonstrated live Docker container deployments handling autoscaled web traffic.
My co-speaker James Gabriele Torzar covered the managed App Engine and Cloud Functions tiers, giving attendees a complete panoramic mental model of modern cloud compute.