plugin-barman-cloud/kubernetes/deployment.yaml
Sai Sridhar 949d8f2de9 fix: change default log level from debug to info in deployment.yaml
The operator deployment had --log-level=debug hardcoded, causing verbose
lifecycle and reconciliation logs in production (e.g. "generated patch",
WAL archiving details). This adds significant noise to log aggregation
systems (Loki, CloudWatch, etc.) without actionable signal.

Change the default to info. Debug logging is still available by patching
the deployment args or setting the flag explicitly.

Part of #917 (global default; Helm value and per-cluster override are
follow-up work).
2026-06-09 10:11:02 +02:00

72 lines
1.7 KiB
YAML

apiVersion: apps/v1
kind: Deployment
metadata:
labels:
app: barman-cloud
name: barman-cloud
spec:
replicas: 1
selector:
matchLabels:
app: barman-cloud
strategy:
type: Recreate
template:
metadata:
labels:
app: barman-cloud
spec:
securityContext:
runAsNonRoot: true
seccompProfile:
type: RuntimeDefault
serviceAccountName: plugin-barman-cloud
containers:
- image: plugin-barman-cloud:latest
name: barman-cloud
ports:
- containerPort: 9090
protocol: TCP
env:
- name: SIDECAR_IMAGE
valueFrom:
secretKeyRef:
key: SIDECAR_IMAGE
name: plugin-barman-cloud
args:
- operator
- --server-cert=/server/tls.crt
- --server-key=/server/tls.key
- --client-cert=/client/tls.crt
- --server-address=:9090
- --leader-elect
- --log-level=info
readinessProbe:
tcpSocket:
port: 9090
initialDelaySeconds: 10
periodSeconds: 10
volumeMounts:
- mountPath: /server
name: server
- mountPath: /client
name: client
resources: {}
securityContext:
allowPrivilegeEscalation: false
capabilities:
drop:
- ALL
readOnlyRootFilesystem: true
runAsGroup: 10001
runAsUser: 10001
seccompProfile:
type: RuntimeDefault
volumes:
- name: server
secret:
secretName: barman-cloud-server-tls
- name: client
secret:
secretName: barman-cloud-client-tls