diff --git a/docs/getting-started/prerequisites/hardware.md b/docs/getting-started/prerequisites/hardware.md index 76cdb6fa83..9c98833c5c 100644 --- a/docs/getting-started/prerequisites/hardware.md +++ b/docs/getting-started/prerequisites/hardware.md @@ -22,12 +22,14 @@ The site controller runs the NICo control plane on a Kubernetes cluster. A minim **Storage layout**: Total local NVMe capacity should be 4 TiB or greater. Mount 1.7 TiB on `/` (root) on the NVMe OS disk (ext4 or xfs) — typical usage is 200–500 GiB. Mount `/var/lib/containerd` and `/var/lib/kubelet` on a separate NVMe data disk (1+ TiB, ext4/xfs, `noatime`). Consider a dedicated `/var/log` if there is heavy logging. Persistent app storage (SAN/NAS, Rook-Ceph) is not required for NICo itself. -### DPUs on Site Controller (Optional) +### DPUs on Site Controller (Required) -DPUs on site controller nodes are optional and site-owned. If DPUs are installed, ensure the following requirements are met: -- Verify the correct DPU power cable has been ordered from the server vendor. +Site-controller nodes must have Bluefield-3 DPUs. Ensure the following requirements are met: +- You have the correct DPU power cable from the server vendor. +- The Bluefield-3's operating mode is DPU mode. NIC mode is not supported. - For BF3 DPUs, verify link speed and optics: BF3 runs at 200 Gb, so match ports to 200 Gb-capable optics, fiber, or DACs. -- A basic onboard NIC for management is sufficient--no extra ConnectX NICs are needed. +- Verify that the DPU can connect to the outside world (curl -I https://www.nvidia.com) +- The DPUs are at the latest supported firmware version: DOCA 2.9.3 and HBN 2.4.3 ## Compute Systems (Managed Hosts) diff --git a/docs/getting-started/quick-start.md b/docs/getting-started/quick-start.md index 99d4350e6e..3e96b1bd3d 100644 --- a/docs/getting-started/quick-start.md +++ b/docs/getting-started/quick-start.md @@ -33,17 +33,23 @@ The cluster must have: ### Site controller node DPU requirements -If your site controller nodes are equipped with DPUs (BlueField NICs), the DPUs must be fully provisioned and configured **before** the Kubernetes cluster is set up. NICo does not provision the site controller nodes' own DPUs — it only manages DPUs on downstream bare-metal hosts after ingestion. +Site controller nodes must be equipped with fully provisioned DPUs (Bluefield-3s) which are configured **before** the Kubernetes cluster is set up. We do not support configuring site controller nodes without DPUs today. NICo does not provision the site controller nodes' own DPUs — it only manages DPUs on downstream bare-metal hosts after ingestion. Specifically, you must complete the following before proceeding: -- Flash the DPU firmware to a supported version using the BlueField Firmware Bundle. -- Configure the DPU operating mode (DPU mode or NIC mode) to match your site controller networking topology. See the [network prerequisites](prerequisites/network.md) for the supported topologies. +- Flash the DPU firmware to the latest supported version using the BlueField Firmware Bundle. Latest supported firmware versions: + + | DOCA | HBN | + | ----- | ----- | + | 2.9.3 | 2.4.3 | + +- Configure the Bluefield-3 device in DPU mode (operating mode). We do not currently support NIC mode. - Ensure the DPU ARM OS is booted and reachable via its management interface. +- Verify that the DPU can connect to the outside world (curl -I https://www.google.com) Refer to the NVIDIA DOCA documentation and the BlueField Firmware Bundle download archive for firmware flashing instructions and supported firmware versions: -[https://developer.nvidia.com/doca-2-9-2-lts-ovs-doca-download-archive?deployment_platform=BlueField&deployment_package=BF-FW-Bundle](https://developer.nvidia.com/doca-2-9-2-lts-ovs-doca-download-archive?deployment_platform=BlueField&deployment_package=BF-FW-Bundle) +[https://developer.nvidia.com/doca-2-9-3-download-archive?deployment_platform=BlueField&deployment_package=BF-FW-Bundle](https://developer.nvidia.com/doca-2-9-3-download-archive?deployment_platform=BlueField&deployment_package=BF-FW-Bundle) ### Required tools (local machine)