Hi all,
I’m having an issue with my Lemmy on K8S that I selfhost. No matter what I do, Pictrs doesn’t want to use my Minio instance. I even dumped the env variables inside the pod, and those seem to be like described in the documentation. Any ideas?
kind: ConfigMap
metadata:
name: pictrs-config
namespace: lemmy
data:
PICTRS__STORE__TYPE: object_storage
PICTRS__STORE__ENDPOINT: http://192.168.1.51:9000
PICTRS__STORE__USE_PATH_STYLE: "true"
PICTRS__STORE__BUCKET_NAME: pict-rs
PICTRS__STORE__REGION: minio
PICTRS__MEDIA__VIDEO_CODEC: vp9
PICTRS__MEDIA__GIF__MAX_WIDTH: "256"
PICTRS__MEDIA__GIF__MAX_HEIGHT: "256"
PICTRS__MEDIA__GIF__MAX_AREA: "65536"
PICTRS__MEDIA__GIF__MAX_FRAME_COUNT: "400"
---
apiVersion: v1
kind: Secret
metadata:
name: pictrs-secret
namespace: lemmy
type: Opaque
stringData:
PICTRS__STORE__ACCESS_KEY: SOMEUSERNAME
PICTRS__STORE__SECRET_KEY: SOMEKEY
PICTRS__API_KEY: SOMESECRETAPIKEY
---
apiVersion: apps/v1
kind: Deployment
metadata:
name: pictrs
namespace: lemmy
spec:
selector:
matchLabels:
app: pictrs
template:
metadata:
labels:
app: pictrs
spec:
containers:
- name: pictrs
image: asonix/pictrs
envFrom:
- configMapRef:
name: pictrs-config
- secretRef:
name: pictrs-secret
volumeMounts:
- name: root
mountPath: "/mnt"
volumes:
- name: root
emptyDir: {}
---
apiVersion: v1
kind: Service
metadata:
name: pictrs-service
namespace: lemmy
spec:
selector:
app: pictrs
ports:
- port: 80
targetPort: 8080
type: ClusterIP
Are you seeing any errors in the logs?
No error logs, based on the logs tho it just ignores the config and uses the filesystem. So predictably once my small config mount fills up (this was before emptyDir), it starts to error out saying no more space on disk. Seemingly this didn’t cause any errors for Lemmy, still it doesn’t feel right :)