Diskpipe with Kubernetes

Use Diskpipe through its CSI adapter, Helm chart and Kubernetes policy resources. This path works on physical or virtual cluster nodes, including cloud clusters.

Read cost control for evaluations before provisioning or moving evaluation data.

Install and validate

• Getting started: the narrated volatile-memory happy path.
• Detailed installation: configuration, validation and troubleshooting.
• Custom kernel build: build the client edge for a kernel outside the supplied matrix.
• Single-region HA checks: graded fault checks with the optional chaos toolbox.

The getting-started memory example is volatile. Follow the guide's durability choices before using storage for data that must survive restart.

Replication and failover

• Cross-region replication.
• Cross-region policy.
• Cross-region failover.
• Tiering.
• Untrusted storage and untrusted hosting.

Day-two operations

Identity and failover operations explains native Kubernetes identity alongside OAuth and IAM, tenant-scoped permissions, approval and teardown checks. Kubernetes credentials authenticate a principal; storage authorization still follows explicit role bindings.

See the Helm chart reference and raw deployment reference for packaging.

Other deployment choices

Start at Diskpipe or explore Diskpipe without Kubernetes.