I deployed kes to kubernetes using the minio operator and noticed that after some time the memory limits are reached for all replicas in the same sequence. This looks like a memory leak to me. Is anything already reported about this?
Additional context
Deployed by minio-operator
Tenant configuration (via kubectl describe tenants.minio.min.io ...):
Kes:
Annotations:
Image: minio/kes:2024-01-11T13-09-29Z
Image Pull Policy: IfNotPresent
Kes Secret:
Name: dc-storage-kes-configuration
Key Name: default-minio-key
Node Selector:
kubernetes.io/arch: arm64
node.kubernetes.io/role: agent
Replicas: 3
Resources:
Limits:
Cpu: 300m
Memory: 400Mi
Requests:
Cpu: 100m
Memory: 100Mi
Security Context:
Fs Group: 1000
Run As Group: 1000
Run As Non Root: true
Run As User: 1000
Name: dc-storage-kes
CreationTimestamp: Tue, 09 Jan 2024 10:28:01 +0100
Selector: v1.min.io/kes=dc-storage-kes
Labels: <none>
Annotations: <none>
Replicas: 3 desired | 3 total
Update Strategy: RollingUpdate
Pods Status: 3 Running / 0 Waiting / 0 Succeeded / 0 Failed
Pod Template:
Containers:
kes:
Image: minio/kes:2024-01-11T13-09-29Z
Port: 7373/TCP
Host Port: 0/TCP
Args:
server
--config=/tmp/kes/server-config.yaml
--auth=off
Limits:
cpu: 300m
memory: 400Mi
Requests:
cpu: 100m
memory: 100Mi
Environment:
MINIO_KES_IDENTITY: XXXX
Mounts:
/tmp/kes from dc-storage-kes (rw)
Volumes:
dc-storage-kes:
Type: Projected (a volume that contains injected data from multiple sources)
SecretName: dc-storage-kes-configuration
SecretOptionalName: <nil>
SecretName: dc-storage-kes-tls
SecretOptionalName: <nil>
Volume Claims: <none>
Events: <none>
Bug describtion
I deployed kes to kubernetes using the minio operator and noticed that after some time the memory limits are reached for all replicas in the same sequence. This looks like a memory leak to me. Is anything already reported about this?
Additional context
Deployed by minio-operator Tenant configuration (via kubectl describe tenants.minio.min.io ...):