Using Proxies and a VPS for Web Scraping and Automation
Learn when to move a scraper from your PC to a VPS, how to size the server, and how to connect it to ARNPROXY.
Guest article by PSB Hosting, adapted by ARNPROXY as part of a content exchange. This article contains a PSB Hosting referral link.
A scraper may run perfectly on your laptop for ten minutes and stop halfway through an overnight job when the laptop sleeps or loses its connection. A virtual private server (VPS) lets the job continue when your computer is off. The proxy determines which IP address the target site sees.
When a home PC becomes a bottleneck
A home computer is fine for short tests. Longer jobs compete with your other apps for CPU and memory, especially when several headless browsers run at once. Sleep, updates and Wi-Fi interruptions can stop a job without warning. Keeping the work on a separate server also makes it easier to check progress remotely and isolate project files from personal ones.
What changes on a VPS
You can schedule jobs with cron or a systemd timer, manage the server remotely and keep the runtime in a separate environment. Choose the vCPU, memory and storage for your workload. Check whether CPU resources are shared or dedicated on the plan you choose.
If the server has a fixed outbound IP, you can allowlist it for proxy products that support IP-based authorization. If you use a username and password instead, keep them out of your scripts and logs.
What the server and the proxy each do
Your code runs on the VPS. When your tool uses the proxy, the target sees the proxy’s outbound IP. With ARNPROXY, datacenter IPv4 and rotating plans suit repeated monitoring; a static ISP proxy lets you keep the same IP across sessions. Residential proxies support city-level targeting for regional checks. Use IPv6 only if the target supports it.
Run automated requests only where permitted and respect the target’s request limits.
Sizing the VPS
Use these configurations as starting points, then measure a representative run before increasing concurrency.
| Workload | vCPU | RAM | NVMe storage |
|---|---|---|---|
| Python or Node.js scrapers, API monitoring | 1–2 | 2–4 GB | 40 GB |
| 5–10 parallel headless browsers with Playwright or Puppeteer | 4 | 8 GB | 80 GB |
| More browsers, queues and databases | 6–8+ | 16 GB+ | 160 GB+ |
Watch memory usage first: heavy swapping can slow down every worker. If the job is CPU-bound, benchmark a faster processor before paying for more cores.
Setup in five steps
- Pick a VPS location and test latency to your proxy gateway and target services.
- Create a non-root user, set up SSH keys and restrict access to unused ports.
- Install your tools, such as Scrapy or Playwright, in a virtual environment or container.
- Configure your ARNPROXY gateway or proxy list. If supported, authorize the VPS’s outbound IP and choose a sticky or rotating session.
- Schedule the job and log errors, response times and successful requests. Review traffic usage and adjust concurrency based on a test run.
About PSB Hosting
PSB Hosting offers VPS locations in the Netherlands, Germany, Finland and the United States. Its plans include options with DDR5 memory, NVMe storage and unmetered traffic, plus ready-to-use images for Docker and Node.js. Check the port speed and resources for your plan. Support is available 24/7.
At PSB Hosting, use the code ARNPROXY for 15% off your first order.
When to make the move
Keep short tests on your PC. For a recurring job that needs to keep running without it, start with a small VPS, confirm that requests use the intended proxy, and scale after measuring CPU, memory and response times.
