The installation and configuration of the PowerFlex SDC driver is required on every MC node in the cluster. One node must be configured at a time. Each node will be paused and drained prior to the installation. After the SDC driver is installed, the host is then rebooted.
Copy the SDC driver to a CSV (one MC node per cluster, preferably node 1)
Copy the Windows version of the PowerFlex SDC driver MSI installer to an S2D volume, accessible to all MC Nodes.
You can copy the SDC driver to an S2D volume (ensuring it is available to all nodes in the cluster) with the WAC, with File Explorer, or with PowerShell.
Ideally starting at node 1, pause and drain the node, install the SDC driver, and verify and reboot the node as per the guidelines described in the following sections.
Suspend and drain a node (all MC nodes)
Before continuing, read the official guidance for Failover cluster maintenance procedures failover cluster maintenance procedures:
Before installing the SDC driver, suspend the node:
Suspend-ClusterNode <NodeName> -DrainValidate DrainStatus is "Completed" and the State of the node is "Paused":
Get-ClusterNode -Name <NodeName> | Select-Object DrainStatus, StateInstall the SDC driver (all MC nodes)
Once the node is successfully paused, perform the following steps to install the SDC driver.
Refer to the official Dell PowerFlex 4.5.x Install and Upgrade guide for full details on installing the SDC driver on a Windows host. See Dell PowerFlex 4.5.x Install and Upgrade Guide | Dell US for additional support
Verify the SDC driver is installed correctly (all MC nodes)
& "C:\Program Files\EMC\scaleio\sdc\bin\drv_cfg.exe" -query_mdmsReboot the node (all MC nodes)
Verify that the node is suspended and drained. See:Failover cluster maintenance procedures, and then reboot the node:
Restart-ComputerOnce the node has restarted, verify the node's health and resume it. Refer to: https://learn.microsoft.com/en-us/azure-stack/hci/manage/maintain-servers for complete instructions.
Resume-ClusterNode -Failback ImmediateConfirm that the node has been resumed (all MC nodes)
Get-ClusterNode -Name <NodeName> | Select-Object FailbackStatus, State, StatusInformationFailbackStatus should be "Completed," State should be "Up," and StatusInformation should be "Normal"
Repeat for every MC node in the cluster, one at a time.