38. Googlebot keeps crawling old Drupal-era image URLs on DH, which return an empty 200 — Medium (DH)
Added 16 Nov 2026. Source: Crawl stats "By file type" exports, 16 Nov 2026 (crawl_stats/2026-11-16/dh-other-file-type/, data 18 Aug – 14 Nov 2026).
What: all 367 sampled "Other file type" fetches on DH are old Drupal image paths under /sites/dh/files/… (article images, gallery images, styles/…?itok= image-style URLs), e.g. /sites/dh/files/article_images/2018/07/15/file710eaag883m1279z05d0153164.jpg. Live check (16 Nov 2026, Googlebot UA): 200, 0 bytes, no content type, no cache header, so every request reaches the origin.
Evidence: DH "Other file type" requests: 17–31K/day from 11 Nov 2026 (after the AMP redirects stopped, #14), i.e. roughly 5–9% of DH's Googlebot requests. Before 11 Nov the same category also held the AMP 301s (63–164K/day).
Cause: tpml-dh app/server/app.js:321-323 @ 47a7e85e: app.use(["/sites", "/sites/:slug"], … res.status(200).send()). An empty 200 tells Google the URLs are alive, so it keeps re-crawling them (also listed under #28).
🛠️ Fix: return 410 Gone for /sites/* (or 301 to the current image on media.assettype.com where a mapping exists, if any of these still earn image-search traffic; check Search Console → Search type: Image, page contains /sites/dh/files/ first). Verify: live curl -sI returns 410; Crawl stats "Other file type" for DH falls toward zero over the following weeks.