Title
Create new category
Edit page index title
Edit category
Edit link
Kubernetes
This guide provides installation and operating instructions for running MetaDefender Cluster on Kubernetes.
MetaDefender Cluster is deployed with two Helm charts.
md-cluster-servicesinstalls the cluster's core services — MetaDefender Cluster Control Center, MetaDefender Cluster Identity Service and MetaDefender Cluster File Storage — together with the PostgreSQL, Redis and RabbitMQ dependencies.md-cluster-instancesinstalls the MetaDefender Cluster Worker instances that perform the actual work: the MetaDefender Core, the MetaDefender Cluster API Gateway and the MetaDefender Cluster Callback Service.
Guide structure
Read the pages in order for a first deployment. Each page is self-contained once the cluster is running.
Page | What it covers |
|---|---|
The two charts, the workloads they create, and the shared ConfigMap and Secret | |
Health probes, rollout behaviour, graceful shutdown and self-healing — reference, not a procedure | |
Prerequisites and the step-by-step install | |
CPU, memory and ephemeral storage sizing | |
How license activation and release work across pod lifecycles | |
Persistent volumes for PostgreSQL and MetaDefender Cluster File Storage | |
Chart values, environment variables, and how to use external infrastructure | |
Upgrading services and instances, rollback, clean removal | |
Clustered PostgreSQL, RabbitMQ, Redis and MetaDefender Cluster File Storage | |
Adding and removing MetaDefender Cluster Workers and MetaDefender Cluster File Storage replicas | |
Symptoms, checks and log collection |
Supported Kubernetes versions
Version | Notes |
|---|---|
1.31.x | Fully supported and validated. |
Earlier versions | Not validated. The charts use only stable APIs ( |
Storage support
Persistent volumes must be backed by block or MetaDefender Cluster File Storage that supports ReadWriteOnce and honours POSIX file ownership.
Do not use Network File System (NFS) for persistent volumes. NFS can corrupt the MetaDefender Cluster File Storage metadata database and the PostgreSQL data directory. Longhorn is a tested alternative for on-cluster storage. Managed cloud block storage (EBS, Azure Disk, Persistent Disk) is also suitable.