I’ve spun up a Kubernetes namespace and populated it with stuff: PVC’s for appmedia and dbdata; deployments and services for broker, db, and webodm; and an ingress for webodm.
All deployments are successful. All containers show reasonable output. I’m able to create the admin account and log into the front end. I manually add localhost:3000 as a processing node (it’s part of the webodm deployment) and it successfully refreshes – I see JSON in the options box, API versions and engine versions just fine.
I can add a task to the first project. When I select photos for upload, a reasonable value for the location is shown in the name field. I have determined that this is a result of a successful retrieval of 64kb of one of the photos, presumably EXIF data, that is sent to nominatim.openstreetmap.org to perform a reverse lookup. So that’s good.
When I review and then approve the task, the image upload fails. The task shows in red “Cannot upload filename, exceeded max retries (10)”. When I open the Javascript console, I see that the upload tasks have all failed with error code 413 Request Entity Too Large returned by nginx.
The webapp container’s access and error logs in /tmp show no attempts to upload. No errors are logged, and no POST to the API for upload is logged either. When I use the docker-compose approach, the access logs show POSTs to the API for upload.
What should I look at next? Thanks in advance for any help.