Next.js Discord

Discord Forum

API route fetch fail when running on dockerized production build image

Answered
stefnto posted this in #help-forum
Open in Discord
Avatar
stefntoOP
I have dockerized my next 14 frontend service and when an api route is called I get this error
⨯ TypeError: fetch failed
2024-06-04 12:21:38     at node:internal/deps/undici/undici:12345:11
2024-06-04 12:21:38     at process.processTicksAndRejections (node:internal/process/task_queues:95:5)
2024-06-04 12:21:38     at async l (/app/.next/server/app/api/auth/route.js:1:3057)
2024-06-04 12:21:38     at async /app/node_modules/next/dist/compiled/next-server/app-route.runtime.prod.js:6:34666
2024-06-04 12:21:38     at async eS.execute (/app/node_modules/next/dist/compiled/next-server/app-route.runtime.prod.js:6:25813)
2024-06-04 12:21:38     at async eS.handle (/app/node_modules/next/dist/compiled/next-server/app-route.runtime.prod.js:6:35920)
2024-06-04 12:21:38     at async doRender (/app/node_modules/next/dist/server/base-server.js:1377:42)
2024-06-04 12:21:38     at async cacheEntry.responseCache.get.routeKind (/app/node_modules/next/dist/server/base-server.js:1599:28)
2024-06-04 12:21:38     at async NextNodeServer.renderToResponseWithComponentsImpl (/app/node_modules/next/dist/server/base-server.js:1507:28)
2024-06-04 12:21:38     at async NextNodeServer.renderPageComponent (/app/node_modules/next/dist/server/base-server.js:1924:24) {
2024-06-04 12:21:38   cause: Error: connect ECONNREFUSED 127.0.0.1:8000
2024-06-04 12:21:38       at TCPConnectWrap.afterConnect [as oncomplete] (node:net:1595:16)
2024-06-04 12:21:38       at TCPConnectWrap.callbackTrampoline (node:internal/async_hooks:130:17) {
2024-06-04 12:21:38     errno: -111,
2024-06-04 12:21:38     code: 'ECONNREFUSED',
2024-06-04 12:21:38     syscall: 'connect',
2024-06-04 12:21:38     address: '127.0.0.1',
2024-06-04 12:21:38     port: 8000
2024-06-04 12:21:38   }
2024-06-04 12:21:38 }
Answered by stefnto
Apparently when running the docker the --network="host" option should be set for the docker and machine running networks to be the same. Sadly this option is available only on linux (WSL supports this as well)
View full answer

2 Replies

Avatar
stefntoOP
Basically its the login client component which calls the api/auth, where in it there is a fetch to the backend where it runs on localhost:8000.
In the backend I've whitelisted localhost, 0.0.0.0 asnd 127.0.0.1.
If I rewrite the backend fetch call to happen inside my Login client component and not inside the api route, I dont get any error, so I suspect there is something going on inside the api route. Anyone had a similar error?
Avatar
stefntoOP
Apparently when running the docker the --network="host" option should be set for the docker and machine running networks to be the same. Sadly this option is available only on linux (WSL supports this as well)
Answer