Blog

Practical field notes for choosing proxy services and operating them reliably. Every article includes a test, worksheet, decision tree, or other artifact you can use, not just prose.

Latest field notes

15 articles

Aug 5, 2026

Akamai 403 During Authorized Automation: What to Record Before Contacting the Site Owner

A screenshot of an Access Denied page gives the site owner little to search. Capture one clean incident, redact it carefully, and send a packet they can correlate with their own logs.

Operations

11 min

Aug 4, 2026

Playwright Proxy Authentication Across Chromium, Firefox, and WebKit: A Tested Compatibility Matrix

We ran one local authenticated proxy fixture through Playwright's three bundled engines. The supported configuration stayed consistent. The failure surfaces did not.

Operations

12 min

Aug 4, 2026

HTTP 200, but the Automation Still Failed: Build a Validator That Catches It

A main-document 200 is a real HTTP success. This field guide adds the page, session, outcome, and budget checks that decide whether your authorized browser job actually worked.

Operations

9 min

Aug 4, 2026

Proxy Price per GB Is Misleading: Calculate Cost per Successful Page

The quoted rate tells you what traffic costs. This calculator shows what each usable page cost in one observed run.

Procurement

9 min

Aug 4, 2026

Residential vs. ISP Proxies for Browser Automation

Residential and ISP can indicate the exit-host architecture. The exact product determines session behavior, allocation, protocol, billing, and replacement rules.

Procurement

10 min

Aug 1, 2026

Playwright ERR_TUNNEL_CONNECTION_FAILED: A Reproducible Checklist

The Chromium error places the failure before a usable proxy tunnel formed. This checklist helps you reproduce the boundary without guessing at the cause.

Operations

7 min

Aug 1, 2026

How to Test a Proxy Provider Before You Commit: A 30-Minute Workload Benchmark

Freeze one authorized workload, run paired controls, and score every started attempt. Thirty minutes can screen a route without pretending to measure a whole provider.

Operations

8 min

Aug 1, 2026

Why Proxy “Success Rate” Numbers Disagree

One set of network events can produce several honest percentages. The number becomes comparable only when its unit, denominator, validator, retries, and taxonomy are visible.

Operations

8 min

Jul 31, 2026

403 vs. 407 vs. 429 vs. Timeout: Which Part of Your Proxy Request Failed?

Start by separating a received HTTP response from a client deadline or network failure. Then trace the exchange that produced it.

Operations

6 min

Jul 31, 2026

Proxy Latency, Origin Latency, or Browser Timeout? How to Measure the Difference

A Playwright timeout is a deadline crossing. Split the request into phases before assigning the delay to a proxy, an origin, or the browser.

Operations

5 min

Jul 30, 2026

Proxy Works in cURL but Fails in Playwright: A Layer-by-Layer Diagnostic

A four-cell test for finding the first place your direct, proxied, command-line, and browser results stop agreeing.

Operations

8 min

Jul 20, 2026

Sticky Proxy Session Keeps Changing IP: Five Causes and a Repeatable Test

A sticky session is a bounded, provider-defined mapping. Record the whole session scope before calling an exit change provider rotation.

Operations

7 min

Jul 20, 2026

How we score: the gate and five scored dimensions

One number can hide a lot. Here is how the six-check gate and five-dimension Proxy Index determine whether a proxy can be scored and ranked.

Methodology

2 min

Jul 20, 2026

What is ASN diversity, and why does it matter for scraping?

Two proxy pools with the same IP count can behave completely differently, and the difference usually lives in the autonomous systems behind them.

Concepts

2 min

Jul 20, 2026

Reading a proxy provider's pool size claim

"70 million IPs" is a marketing number. The number that determines your success rate is much smaller, and providers rarely print it.

Procurement

2 min