fix(workers): Fix the leaky browser instances in workers during development

3 files changed, 46 insertions, 29 deletions

diff --git a/apps/workers/crawlerWorker.ts b/apps/workers/crawlerWorker.ts
index ecd8d146..eb4a0697 100644
--- a/apps/workers/crawlerWorker.ts
+++ b/apps/workers/crawlerWorker.ts
@@ -1,40 +1,36 @@
-import logger from "@hoarder/shared/logger";
-import {
-  LinkCrawlerQueue,
-  OpenAIQueue,
-  SearchIndexingQueue,
-  ZCrawlLinkRequest,
-  queueConnectionDetails,
-  zCrawlLinkRequestSchema,
-} from "@hoarder/shared/queues";
+import assert from "assert";
+import { Readability } from "@mozilla/readability";
+import { Mutex } from "async-mutex";
+import { Job, Worker } from "bullmq";
 import DOMPurify from "dompurify";
+import { eq } from "drizzle-orm";
+import { isShuttingDown, shutdownPromise } from "exit";
 import { JSDOM } from "jsdom";
-
-import { Worker } from "bullmq";
-import { Job } from "bullmq";
-
-import { db } from "@hoarder/db";
-
-import { Browser } from "puppeteer";
-import puppeteer from "puppeteer-extra";
-import StealthPlugin from "puppeteer-extra-plugin-stealth";
-import AdblockerPlugin from "puppeteer-extra-plugin-adblocker";
-
 import metascraper from "metascraper";
-
 import metascraperDescription from "metascraper-description";
 import metascraperImage from "metascraper-image";
 import metascraperLogo from "metascraper-logo-favicon";
+import metascraperReadability from "metascraper-readability";
 import metascraperTitle from "metascraper-title";
-import metascraperUrl from "metascraper-url";
 import metascraperTwitter from "metascraper-twitter";
-import metascraperReadability from "metascraper-readability";
-import { Mutex } from "async-mutex";
-import assert from "assert";
-import serverConfig from "@hoarder/shared/config";
+import metascraperUrl from "metascraper-url";
+import { Browser } from "puppeteer";
+import puppeteer from "puppeteer-extra";
+import AdblockerPlugin from "puppeteer-extra-plugin-adblocker";
+import StealthPlugin from "puppeteer-extra-plugin-stealth";
+
+import { db } from "@hoarder/db";
 import { bookmarkLinks } from "@hoarder/db/schema";
-import { eq } from "drizzle-orm";
-import { Readability } from "@mozilla/readability";
+import serverConfig from "@hoarder/shared/config";
+import logger from "@hoarder/shared/logger";
+import {
+  LinkCrawlerQueue,
+  OpenAIQueue,
+  queueConnectionDetails,
+  SearchIndexingQueue,
+  ZCrawlLinkRequest,
+  zCrawlLinkRequestSchema,
+} from "@hoarder/shared/queues";
 
 const metascraperParser = metascraper([
   metascraperReadability(),
@@ -60,6 +56,12 @@ async function launchBrowser() {
       userDataDir: serverConfig.crawler.browserUserDataDir,
     });
     browser.on("disconnected", async () => {
+      if (isShuttingDown) {
+        logger.info(
+          "The puppeteer browser got disconnected. But we're shutting down so won't restart it.",
+        );
+        return;
+      }
       logger.info(
         "The puppeteer browser got disconnected. Will attempt to launch it again.",
       );
diff --git a/apps/workers/exit.ts b/apps/workers/exit.ts
new file mode 100644
index 00000000..156cf2c6
--- /dev/null
+++ b/apps/workers/exit.ts
@@ -0,0 +1,11 @@
+import logger from "@hoarder/shared/logger";
+
+export let isShuttingDown = false;
+
+export const shutdownPromise = new Promise((resolve) => {
+  process.on('SIGTERM', () => {
+    logger.info("Received SIGTERM, shutting down ...");
+    isShuttingDown = true;
+    resolve("");
+  });
+});
diff --git a/apps/workers/index.ts b/apps/workers/index.ts
index 5b6b62d5..4db524ef 100644
--- a/apps/workers/index.ts
+++ b/apps/workers/index.ts
@@ -2,6 +2,7 @@ import "dotenv/config";
 import { CrawlerWorker } from "./crawlerWorker";
 import { OpenAiWorker } from "./openaiWorker";
 import { SearchIndexingWorker } from "./searchWorker";
+import { shutdownPromise } from "./exit";
 
 async function main() {
   const [crawler, openai, search] = [
@@ -10,7 +11,10 @@ async function main() {
     await SearchIndexingWorker.build(),
   ];
 
-  await Promise.all([crawler.run(), openai.run(), search.run()]);
+  await Promise.any([
+    Promise.all([crawler.run(), openai.run(), search.run()]),
+    shutdownPromise,
+  ]);
 }
 
 main();