From 322996347bf7788e4e773949671f38d065d6143c Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Gerg=C5=91=20M=C3=B3ricz?= Date: Sun, 14 Sep 2025 13:24:57 +0200 Subject: [PATCH] fix(scrape-worker): cap sitemaps per crawl at 20 to avoid queue flooding --- apps/api/src/services/worker/scrape-worker.ts | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/apps/api/src/services/worker/scrape-worker.ts b/apps/api/src/services/worker/scrape-worker.ts index 02efe206c..1ddcf1ce1 100644 --- a/apps/api/src/services/worker/scrape-worker.ts +++ b/apps/api/src/services/worker/scrape-worker.ts @@ -713,6 +713,16 @@ async function addKickoffSitemapJob( sc: StoredCrawl, logger: Logger, ) { + // TEMP: max 20 sitemaps per crawl + if ( + (await redisEvictConnection.scard( + "crawl:" + sourceJob.data.crawl_id + ":sitemaps", + )) >= 20 + ) { + logger.debug("Sitemap limit reached, skipping...", { sitemap: sitemapUrl }); + return; + } + const sitemapLocked = (await redisEvictConnection.sadd( "crawl:" + sourceJob.data.crawl_id + ":sitemaps",