Julien Veyssier

@sarunaskas > Could you guide me on how to check it? If you know the connection bandwidth on your server side, you can estimate how much data can be downloaded...

I think the problem here is that we reach the client timeout. A backoff plugin won't solve that. I tried to remove the timeout in the app. Let's see if...

The "entity" which downloads the files is your Nextcloud server. To do so, it uses an HTTP client (GuzzleHttp). When this client makes requests (ask things to the API, download...

Ok we're almost there. Silly mistake. The tmp files were left there so they were only deleted when the background job had finished. They are now deleted directly after having...

Btw, it would be much better to avoid storing tmp files. It turn out Guzzle is able to download in a stream (so the file could directly be written in...

Do you know if one of your drive files is bigger than 3 GB?

Wow this is strange then. I don't know what is taking so much space. What the app does is pretty straightforward: For each drive file: * download it in a...

> it went well after I mapped /tmp to my large external storage Great! We did it! It would still be better to avoid this full storage issue by solving...

I finally found the problem and managed to directly download to the target file instead of using a temp file. If one day you find time and you feel like...

Neverending story... After a deep search in Nextcloud File/Node implementation, the only reason that could cause `fopen` to return `null` (that what we see in your logs) would be that...