Skip to content

Scraping Without Blocks · Appendix

Appendix

Reference pages for the crawl desk: a status code table, a pre-flight checklist, a debugging order, a glossary and links to ProxyLook tools.

  • HTTP status quick reference
  • Pre-flight checklist
  • Debugging order by layer
  • Glossary and further reading

4 min read · page 44 of the PDF

A crawl that returned 40,000 records yesterday returns 6,000 today, and the dashboard is red. These pages are for that moment. They condense the book into tables and lists that can be read in a minute.

HTTP status quick reference

The table gives the usual meaning of each result for a scraper and the first reaction. Chapter 3 turns it into code.

ResultUsual meaningFirst reaction
200Served, or a soft blockValidate the body before storing
301, 302Moved, or sent to a challenge or loginCheck the final URL
304Unchanged since the last fetchKeep the stored copy
403Identity refusedNew session; escalate if it repeats
404, 410Page is goneRecord it; do not retry
407Proxy rejected the credentialsFix the configuration
429Rate limit reachedWait as told; slow the domain
500Origin errorBack off and retry
502, 503, 504Overloaded origin, gateway fault or challengeBack off; inspect the body
Timeout or resetDead exit or a deliberate stallNew session, then retry
Common results, what each usually means for a scraper, and the first reaction.

Pre-flight checklist

Run through this list before pointing a crawler at a new target.

  • Confirm the data is public, and review personal data and the site's terms.
  • Read robots.txt and note any crawl delay.
  • Probe one page with an HTTP client, then a browser, and assign a tier.
  • Look for the value in raw HTML or embedded JSON before using a browser.
  • Pick the cheapest proxy type that clears the tier.
  • Verify the exit country and match the language header.
  • Set timeouts, an attempt cap and a per-domain rate limit.
  • Define the validation marker and minimum size for a good page.
  • Set a daily traffic budget with an alert.
  • Check that logs record status, final URL, route and session ID.

Debugging order

When success drops, work from the cheapest check to the most expensive. The order follows the detection layers from chapter 1, with configuration first and data last.

1 · ConfigurationCredentials, gateway address, country parameter, remaining balance.
2 · Exit IPType, country and network of the address the target saw.
3 · RateRequests per domain per minute, and any Retry-After header.
4 · Client signatureHTTP version and headers compared with a current browser.
5 · Browser runtimeLocale, time zone, automation flags, blocked resources.
6 · Session behaviorPacing, page order and session length.
7 · DataBody validation and batch checks for soft blocks.
The order in which to check a failing crawl, cheapest first.

Glossary

Short definitions of terms used in this book. The ProxyLook glossary covers proxy vocabulary more widely.

TermMeaning
ASNThe number of the network that owns an IP range
BackoffWaiting longer after each failed attempt
Browser contextAn isolated browser profile with its own cookies and proxy
Challenge pageAn interstitial asking for proof of a browser or person
Dead-letter queueA holding area for jobs that used up their retries
Exit IPThe address the target sees
FingerprintClient traits that together identify software or a device
Idempotent jobA job that stores the same result when run twice
JA3, JA4Methods that hash a TLS ClientHello into a fingerprint
JitterRandomness added to waits so workers do not retry together
p95 latencyThe time under which 95% of attempts finish
Rotating proxyA gateway that assigns a new exit IP per request
Routing ladderRoutes ordered from cheapest to most capable
Scraper APIA service that fetches a URL and returns the page
Soft blockA 200 response with altered or missing data
Sticky sessionOne exit IP held for a sequence of requests
UnblockerA scraper API exposed as a proxy endpoint
Terms used in this book.

ProxyLook tools and further reading

ProxyLook is an independent directory that lists more than 200 proxy providers. Its free tools cover three of the checks above.

Two companion books sit beside this one. The Proxy Playbook is the flagship. It covers proxy types, protocols and basic setup. The Multi-Account Handbook is for teams that run several accounts. Multi-accounting can breach a platform's terms of service, and the operator carries that risk.

Key takeaways

  • Read the status code as a clue to the layer, then validate the body anyway.
  • Run the pre-flight list before the first request to a new target.
  • Debug from configuration upward and change one variable per test.
  • Keep the legal questions about personal data, logins and terms separate from the technical ones.

Prefer the designed PDF? Scraping Without Blocks is 49 pages, free, with no sign-up.

Download the PDF