<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>Infrastructure on Solvinc — AI's Uncharted Edge</title><link>https://blog.solvinc.io/tags/infrastructure/</link><description>Recent content in Infrastructure on Solvinc — AI's Uncharted Edge</description><generator>Hugo -- 0.139.0</generator><language>en-us</language><lastBuildDate>Tue, 17 Nov 2026 00:00:00 +0700</lastBuildDate><atom:link href="https://blog.solvinc.io/tags/infrastructure/index.xml" rel="self" type="application/rss+xml"/><item><title>You Already Shipped a Search Engine and Didn't Charge for It: SQLite FTS5 Over Scraped Data, Tested</title><link>https://blog.solvinc.io/posts/fts5-search-engine-over-scraped-data-sqlite/</link><pubDate>Tue, 17 Nov 2026 00:00:00 +0700</pubDate><guid>https://blog.solvinc.io/posts/fts5-search-engine-over-scraped-data-sqlite/</guid><description>Your scraper already collects searchable data. Ship it as a product: SQLite FTS5 with bm25 ranking, triggers, and zero infrastructure, tested at 50k rows.</description></item><item><title>The Free Discount Hiding in Your Scraper: Conditional GET, 304s, and the Proxy Bill You're Paying for Nothing</title><link>https://blog.solvinc.io/posts/conditional-get-304-monitoring-cheaper-proxy-bills/</link><pubDate>Mon, 16 Nov 2026 00:00:00 +0700</pubDate><guid>https://blog.solvinc.io/posts/conditional-get-304-monitoring-cheaper-proxy-bills/</guid><description>Stock HTTP clients never send If-None-Match, so your monitors re-download unchanged pages at proxy rates. Tested validator store, A/B harness, and a challenge audit for lying ETags.</description></item><item><title>Your Playwright Fleet Loses 92 MB per Task — a Supervisor That Fixes Leaks, Crashes, and Zombie Browsers</title><link>https://blog.solvinc.io/posts/browser-fleet-supervisor-playwright-crash-recovery/</link><pubDate>Fri, 13 Nov 2026 00:00:00 +0700</pubDate><guid>https://blog.solvinc.io/posts/browser-fleet-supervisor-playwright-crash-recovery/</guid><description>Leaked Playwright contexts cost ~93 MB of RSS each on Chromium 153. A tested fleet supervisor: crash recovery by liveness probing, deadline kills, and a zombie reaper.</description></item><item><title>The Other Side of the Anti-Bot Arms Race: A Tripwire Layer That Fingerprints Bots by What They Can't Resist (Tested Python)</title><link>https://blog.solvinc.io/posts/scraper-tripwire-layer-seed-fields-trap-paths-defender-side/</link><pubDate>Thu, 12 Nov 2026 00:00:00 +0700</pubDate><guid>https://blog.solvinc.io/posts/scraper-tripwire-layer-seed-fields-trap-paths-defender-side/</guid><description>Trap paths, seed fields and keyless decoys: a defender-side bot tripwire that catches scrapers and form spam at the app layer. Tested, 37 assertions green.</description></item><item><title>Your Scraper Needs Japan at 3 A.M. Self-Hosted Geo Egress With WireGuard and HAProxy</title><link>https://blog.solvinc.io/posts/self-hosted-geo-egress-pool-wireguard-haproxy/</link><pubDate>Wed, 11 Nov 2026 04:08:00 +0700</pubDate><guid>https://blog.solvinc.io/posts/self-hosted-geo-egress-pool-wireguard-haproxy/</guid><description>WireGuard + HAProxy geo egress pool on cheap VPSes, instead of $5.88/GB residential proxies. Tested configs, tested asyncio router, real per-GB math.</description></item><item><title>Your Scraper Logged In Once and Never Again: A Session Pool That Survives Login Walls, Challenges, and Token Rot</title><link>https://blog.solvinc.io/posts/authenticated-session-pool-login-wall-scraping-python/</link><pubDate>Tue, 10 Nov 2026 04:00:00 +0700</pubDate><guid>https://blog.solvinc.io/posts/authenticated-session-pool-login-wall-scraping-python/</guid><description>A tested authenticated session pool for login-wall scraping — EWMA health scoring, challenge quarantine, auto-replenishment. Python, SQLite, zero API keys.</description></item><item><title>Stop Burning Proxy Credits: A Health-Scoring Pool That Auto-Quarantines Burned IPs Before They Cost You</title><link>https://blog.solvinc.io/posts/proxy-pool-health-scoring-auto-quarantine-sqlite-python/</link><pubDate>Mon, 09 Nov 2026 15:08:00 +0700</pubDate><guid>https://blog.solvinc.io/posts/proxy-pool-health-scoring-auto-quarantine-sqlite-python/</guid><description>A proxy pool that scores every IP by EWMA block-rate, quarantines burned IPs, and steers traffic to healthy ones. Tested Python, SQLite, zero API keys.</description></item><item><title>Your Webhook Handler Loses Money on Every Deploy: The SQLite Inbox That Makes the Provider's Retry Loop Do Your Queuing</title><link>https://blog.solvinc.io/posts/durable-webhook-ingestion-sqlite-provider-retries-python/</link><pubDate>Mon, 09 Nov 2026 00:00:00 +0700</pubDate><guid>https://blog.solvinc.io/posts/durable-webhook-ingestion-sqlite-provider-retries-python/</guid><description>Webhook handlers silently drop events on every deploy. A tested SQLite inbox turns the provider&amp;#39;s 3-day retry loop into your free queue.</description></item><item><title>Your Scraper Didn't Crash. It Started Lying: A Drift Sentinel That Catches Silent Breakage (Tested Python + SQLite)</title><link>https://blog.solvinc.io/posts/scraper-drift-detection-silent-breakage-sentinel/</link><pubDate>Sun, 08 Nov 2026 00:00:00 +0700</pubDate><guid>https://blog.solvinc.io/posts/scraper-drift-detection-silent-breakage-sentinel/</guid><description>Scrapers fail quietly: empty runs, null rates creeping up, layout drift. A tested DB-level drift sentinel that catches all three before your data customers do.</description></item><item><title>I Deleted a Live Production Database Mid-Write and Lost 2 Rows: The Durability Runbook Every Automation Product Skips</title><link>https://blog.solvinc.io/posts/sqlite-wal-litestream-disaster-recovery-automation-products/</link><pubDate>Sat, 07 Nov 2026 00:00:00 +0700</pubDate><guid>https://blog.solvinc.io/posts/sqlite-wal-litestream-disaster-recovery-automation-products/</guid><description>A kill-window test, a SIGKILL disaster drill, and a Litestream restore: I deleted a live SQLite DB mid-write and got 298 of 300 rows back. Full runbook.</description></item><item><title>The Crawl Frontier: The Scraping Infrastructure Nobody Talks About (Tested Python + SQLite)</title><link>https://blog.solvinc.io/posts/crawl-frontier-politeness-python-sqlite/</link><pubDate>Sat, 07 Nov 2026 00:00:00 +0700</pubDate><guid>https://blog.solvinc.io/posts/crawl-frontier-politeness-python-sqlite/</guid><description>The queue, not the HTTP layer, is where scrapers die. A tested two-level frontier with per-domain politeness, robots.txt, and crash recovery in ~180 lines.</description></item></channel></rss>