Unable to load Worlds on Raspberrypi

[Michi9409]

Hey everyone,
I recently set up Foundry on my Raspberypi 4 (Raspbian V11) and everything went smoothly until I tried to load a world. In the dev tools of my browser I found the error shown on the picture I attached. Can anyone help me resolve that problem?

I have docker version 20.10.5+dfsg1 and docker-compose version 1.25.0 installed.

Here is the content of the docker-compose.yml I am using:

---

version: "3.3"

services:

foundry:
image: felddy/foundryvtt:release
hostname: my-domain
restart: "unless-stopped"
volumes:
- type: bind
source: /path/to/foundrydata
target: /data
environment:
- FOUNDRY_PASSWORD=my-foundry-password
- FOUNDRY_USERNAME=my-foundry-username
- FOUNDRY_ADMIN_KEY=my-admin-key

 ports:
  - target: 30000
    published: 30000
    protocol: tcp

Thanks

[felddy]

Are you accessing the page through a proxy? My guess is that something between you and the application is blocking the web sockets.

See:

• Blank Screen, no admin login box #278 (comment)
• License page empty on launch #118

[Michi9409]

I don't know if that rules that out, but if I reach the pi with its local IP address in my local network, I get the same error. And I did not intentionally set up any proxies.

[felddy]

Is the node process running out of memory? Do you see anything suspicious in the container logs? I see similar reports to yours on Discord that were the result of an out-of-memory condition.

See:

• https://discord.com/channels/170995199584108546/486930822465716249/1159246208209006804

[Michi9409]

I don't know how much memory is normally allocated to Foundry, but it seemed, that the Pi itself did not run out of memory. But I found something in the logs, apparently there is a segmentation fault occuring:

FoundryVTT | 2023-10-12 17:08:10 | [info] Software license verification succeeded
FoundryVTT | 2023-10-12 17:08:10 | [info] Server started and listening on port 30000
FoundryVTT | 2023-10-12 17:08:16 | [info] Created client session
FoundryVTT | 2023-10-12 17:08:17 | [info] Created client session
FoundryVTT | 2023-10-12 17:08:29 | [info] Administrator authentication successful for session
FoundryVTT | 2023-10-12 17:08:33 | [info] Saved application configuration options to /data/Config/options.json
Segmentation fault (core dumped)
Launcher | 2023-10-12 17:08:40 | [error] Node process exited with code 139
Entrypoint | 2023-10-12 17:08:43 | [info] Starting felddy/foundryvtt container v11.311.0
Entrypoint | 2023-10-12 17:08:43 | [info] Foundry Virtual Tabletop 11.311 is installed.
Entrypoint | 2023-10-12 17:08:43 | [info] Not modifying existing installation license key.
Entrypoint | 2023-10-12 17:08:43 | [info] Setting data directory permissions.
Entrypoint | 2023-10-12 17:08:43 | [info] Starting launcher with uid:gid as foundry:foundry.
Launcher | 2023-10-12 17:08:43 | [info] Generating options.json file.
Launcher | 2023-10-12 17:08:43 | [info] Setting 'Admin Access Key'.
Launcher | 2023-10-12 17:08:44 | [info] Starting Foundry Virtual Tabletop.
FoundryVTT | 2023-10-12 17:08:48 | [info] Running on Node.js - Version 18.18.0
FoundryVTT | 2023-10-12 17:08:48 | [info] Foundry Virtual Tabletop - Version 11 Build 311
FoundryVTT | 2023-10-12 17:08:48 | [info] User Data Directory - "/data"
FoundryVTT | 2023-10-12 17:08:48 | [info] Application Options:
{
"awsConfig": null,
"compressSocket": false,
"compressStatic": false,
"fullscreen": false,
"hostname": null,
"hotReload": false,
"language": "en.core",
"localHostname": null,
"passwordSalt": null,
"port": 30000,
"protocol": null,
"proxyPort": null,
"proxySSL": false,
"routePrefix": null,
"sslCert": null,
"sslKey": null,
"updateChannel": "stable",
"upnp": false,
"upnpLeaseDuration": null,
"world": null,
"adminPassword": "••••••••••••",
"cssTheme": "foundry",
"deleteNEDB": false,
"serviceConfig": null
}

[felddy]

This looks like a memory issue. Is the world your are loading very large? Can you try creating a new blank world and see if that also causes a segfault?

[Michi9409]

The test world I created in the screenshot is actually a blank world with no contents. I noticed, when running

$ sudo docker stats

I noticed in the column MEM USAGE / LIMIT, both Mem usage and limit is at all times 0B, as well as MEM % is at 0%. Did I forget to allocate memory to the container? And if so, how do I specify that?

[Michi9409]

I solved the problem. Apparently I had the 32 bit version of Raspbian installed, so I installed the 64 bit version, set everything up as before and it worked flawlessly. Thanks for the help!

[felddy]

I'm glad you figured it out. I appologize for not linking to the existing issue earlier, but it seemed that you were crashing much later than others that had the 32-bit OS installed.

See:

• Node segfault (error 139) in container release v11.x.x #697