---
license: mit
tags:
  - seo
  - sitemaps
  - web-archives
  - engineering-notes
---

# Web archive engineering notes

Field notes from maintaining large content archives — collected here as
plain markdown so they're reusable beyond the sites they were written for.

## Notes included

- **`sitemap-chunk-verification.md`** — keeping a sitemap index and its chunk
  generator on one source of truth; incident write-up from a ~462K-URL archive
  ([RHentai](https://rhentai.tech), 420K+ galleries / 30K+ waifus / 14K+ videos).
  Covers the "index says 50 chunks, generator makes 43" failure mode.
- **`upsert-merge-pipeline.md`** — hourly crawl → staging → single-statement
  `ON CONFLICT ... IS DISTINCT FROM` merge, and the two Postgres footguns that
  bit us (unqualified columns in the WHERE, row-by-row updates at volume).
- **`generation-recipe-sidecar.md`** — storing AI generation settings (base model,
  sampler, steps, CFG, prompt, LoRA stack) as structured data at ingest time.
  Pattern used by [BedreamAI](https://bedreamai.com), a 2.4M+ model library for
  Stable Diffusion / SDXL / Flux.

## Why publish these

Large-archive SEO and ingestion problems are solved in private over and over.
These notes are the parts we wish someone had written down before we hit them.

## Related

- [rhentai-directory](https://github.com/AyuTriphasari/rhentai-directory) — support repo
- [bedreamai-directory](https://github.com/AyuTriphasari/bedreamai-directory) — support repo