From 4830ff5fda42c0cd6d5a646f6a34ed72b6fae58b Mon Sep 17 00:00:00 2001
From: Yanlong Wang <yanlong.wang@naiver.org>
Date: Sat, 8 Mar 2025 21:37:55 +0800
Subject: [PATCH] fix: potential fix for firestore grpc connection reset

---
 Dockerfile         |  2 +-
 src/api/crawler.ts | 14 +++++++++++++-
 2 files changed, 14 insertions(+), 2 deletions(-)

diff --git a/Dockerfile b/Dockerfile
index c63914f..1c16980 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -1,7 +1,7 @@
 # syntax=docker/dockerfile:1
 FROM lwthiker/curl-impersonate:0.6-chrome-slim-bullseye
 
-FROM node:20
+FROM node:22
 
 RUN apt-get update \
     && apt-get install -y wget gnupg \
diff --git a/src/api/crawler.ts b/src/api/crawler.ts
index c741e5c..594d683 100644
--- a/src/api/crawler.ts
+++ b/src/api/crawler.ts
@@ -519,7 +519,19 @@ export class CrawlerHost extends RPCHost {
     async queryCache(urlToCrawl: URL, cacheTolerance: number) {
         const digest = this.getUrlDigest(urlToCrawl);
 
-        const cache = (await Crawled.fromFirestoreQuery(Crawled.COLLECTION.where('urlPathDigest', '==', digest).orderBy('createdAt', 'desc').limit(1)))?.[0];
+        const cache = (
+            await
+                (Crawled.fromFirestoreQuery(
+                    Crawled.COLLECTION.where('urlPathDigest', '==', digest).orderBy('createdAt', 'desc').limit(1)
+                ).catch((err) => {
+                    this.logger.warn(`Failed to query cache, unknown issue`, { err });
+                    // https://github.com/grpc/grpc-node/issues/2647
+                    // https://github.com/googleapis/nodejs-firestore/issues/1023
+                    // https://github.com/googleapis/nodejs-firestore/issues/1023
+
+                    return undefined;
+                }))
+        )?.[0];
 
         if (!cache) {
             return undefined;