Closed MaximUltimatum closed 7 months ago
On the chance that google has lead you here and you are facing a similar issue.... don't comment out lines in a helm chart when debugging something... and then forget to uncomment them.
https://github.com/MaximUltimatum/kube-homelab/blob/master/helm-configs/nextcloud-values.yaml#L28
This was the problem nginx.ingress.kubernetes.io/proxy-body-size: 4G
was commented out
This was the problem nginx.ingress.kubernetes.io/proxy-body-size: 4G was commented out
Yep, that's how I solved it too a while back. Good sleuthing!
Describe your Issue
When attempting to upload files over 1M, the client (desktop windows, desktop linux, android, or directly from web GUI) fails with 413 file too large. I re-enabled debug logging in my helm chart. However, I wasn't seeing the upload failures correspond to any log messages in the nextcloud container. However, when I checked the ingress nginx logs, I am seeing the below error
Logs and Errors
From the ingress-nginx logs
I have checked the corresponding timeframe with debug logging on in the nextcloud container - it registers that there was a GET request for the page when I loaded it, but nothing about the failure to upload the file.
Attempts to solve the issue I've made so far
I've read through this post on the nextcloud forums and [https://github.com/nextcloud/docker/issues/762#issuecomment-504225433] this github issue which appear to be very similar symptoms.
I think my biggest problem is I'm not sure how to properly configure the accepted client size on the ingress-nginx (reverse-proxy?) ingress controller. I'm a little shaky on the interaction between the ingress nginx stuff that helm sets up and the ingress-controller, that (I believe?) I configured on my own. For avoidance of confusion, I'll post the ingress setup directly instead of trying to describe it and risk mistating something.
What I suspect is happening
Typing this all out, I'm wondering if, since the logs don't show up in the nextcloud-pod log and the logs do show up in the pod managed by the nginx-controller-deployment I've incorrectly set up my ingress-controller. I've tried exec-ing into the ingress-controller container directly and adding the
client_max_body_size 10G;
andclient_body_buffer_size 400M;
directly into the nginx.conf file. That didn't work, but I'm wondering if my attempts to restart nginx after editing the file didn't actually restart it. If this is the issue, I don't think it's necessarily related to the nextcloud-helm chart (and more my own lack of understanding of ingress controllers).Describe your Environment
Kubernetes distribution: Vanilla k8s deployed via kubespray
Helm Version (or App that manages helm):
Helm Chart Version:
values.yaml
: I have a testing branch with current values file hereAdditional context, if any
Apologies for submitting another issue so soon. I don't know if I'm off the happy path or just good at finding/creating problems :). Thank you for your help in advance!