How to Size a VPS: CPU, RAM and Disk by Workload

"How many cores and how much RAM do I need?" is the most common pre-sales question. Undersize and your site lags; oversize and you waste money. This guide gives reference configurations per workload and shows how to verify after launch.

Starting points for common workloads

• Personal blog / company site: 1 core, 1–2 GB. Mostly-static sites put almost no pressure on CPU;
• WordPress / online store: start at 2 cores / 4 GB. PHP and MySQL are both memory-hungry — below 4 GB the database becomes the bottleneck;
• Node.js / Python backends: 2 cores, 2–4 GB depending on concurrency;
• Multiple Docker services: budget 256 MB–1 GB per container, then leave 1 GB headroom;
• Dedicated database: the more RAM the better — start at 4 GB with an NVMe disk;
• Remote desktop / multi-tab browser automation: each Chrome tab takes roughly 200–500 MB; heavy usage calls for 8 GB or more.

Which matters most: CPU, RAM or disk?

In most scenarios the bottleneck order is RAM > disk IO > CPU. Running out of RAM triggers the OOM killer or heavy swapping — the server feels frozen. A slow disk feels like slow database queries and page loads. Insufficient CPU feels like slowdowns at peak hours, but the site still works. On a tight budget, prioritize RAM and an NVMe disk.

Verifying after launch

Log in via SSH and watch three commands: free -h for memory (if available stays under 200 MB, upgrade); df -h for disk usage (clean up or expand past 80%); top for CPU load (a load average consistently above your core count means CPU pressure).

00Shark plans support in-place upgrades — start small and scale based on real monitoring data instead of over-buying up front.