Retry middleware loops to max
Likely causes
- - 407/timeout classified as retryable
- - Wrong credentials in env
Verified remedy
Fix the 407 or timeout root cause first; retries cannot repair auth.
Proxy troubleshooting
Ranked causes, ordered diagnostics, and a verified remedy for scrapy retrying all requests - ending with a final connection check.
Quick answer
Fix: Scrapy retries every request through the proxy. Confirm endpoint values, test with curl through the endpoint, apply the matched remedy, then re-verify the exit IP is your assigned endpoint.
Ranked causes, ordered diagnostics, and a verified remedy for scrapy retrying all requests - ending with a final connection check.
Fix: Scrapy retries every request through the proxy. Confirm endpoint values, test with curl through the endpoint, apply the matched remedy, then re-verify the exit IP is your assigned endpoint.
Likely causes
Verified remedy
Fix the 407 or timeout root cause first; retries cannot repair auth.
Likely causes
Verified remedy
Lower CONCURRENT_REQUESTS to 1-2, confirm stability, then increase in steps.
The original action completes and the exit IP matches the endpoint.
If curl through the endpoint also fails, the problem is endpoint-level: check dashboard status or swap to a spare endpoint.
Facts last reviewed 2026-09-01.
RETRY_TIMES and bad-status handling matter less than fixing the underlying proxy failure the retries keep repeating.
Work the diagnosis tree top to bottom: confirm endpoint values, test with curl through the endpoint, apply the remedy, then run the final connection check.
Recurring proxy failures usually trace to an unstable mapping - rotating endpoints or shared addresses. A dedicated static endpoint per profile makes behavior predictable.
Buy self-serve now, or talk with us about a larger agency inventory.