Installing a compute node in a chassis

Use this information to install a compute node in a Lenovo Flex System chassis.

Before you install the compute node in a chassis, read Safety and Installation guidelines.

Statement 21
Graphic illustrating the Safety statement 1 symbol

CAUTION:
Hazardous energy is present when the compute node is connected to the power source. Always replace the compute node cover before installing the compute node.

If you are installing a compute node model that does not have an integrated Ethernet controller, you must install a network interface adapter before you install the compute node in the chassis for management network communication. For a list of supported optional devices for the compute node, see the Lenovo ServerProven website.

To install the compute node in a chassis, complete the following steps.
compute node installation

  1. Select the node bay.
    Notes:
    1. If you are reinstalling a compute node that you removed, you must install it in the same node bay from which you removed it. Some compute node configuration information and update options are established according to node bay number. Reinstalling a compute node into a different node bay can have unintended consequences. If you reinstall the compute node into a different node bay, you might have to reconfigure the compute node.
    2. To maintain proper system cooling, do not operate the Lenovo Flex System chassis without a compute node or node bay filler in each node bay.
    3. If you plan to configure nodes to form a multi-node complex, the lowest node bay ID is the primary node in the scalable partition. The other nodes are the secondary nodes.
    4. If you purchased a multi-node complex and you received each compute node packaged individually, each compute node has a label on the top of the bezel identifying the primary node and the secondary nodes as defined at the factory during testing. Install the nodes in the chassis in the order indicated on the label.
    5. If you plan to configure a scalable partition, you must install and configure compute nodes in the configuration at the same time.
  2. Make sure that the front handles on the compute node are in the open position.
  3. Slide the compute node into the node bay until it stops.
  4. Push the front handles on the front of the compute node to the closed position.
    Note: After the compute node is installed, the IMM2 in the compute node initializes and synchronizes with the Chassis Management Module. This process takes approximately 90 seconds to complete. The power LED flashes rapidly, and the power button on the compute node does not respond until this process is complete.
  5. Optionally, install the SMP expansion connector (see Installing a scalability and performance connector).
  6. Turn on the compute node (see Turning on the compute node for instructions).
  7. Make sure that the power LED on the compute node control panel is lit continuously, indicating that the compute node is receiving power and is turned on.
  8. If you have other compute nodes to install, do so now.
  9. You can place identifying information on the label tabs that are accessible from the front of the compute node.

If this is the initial installation of the compute node in the chassis, you must configure the compute node through the Setup utility and install the compute node operating system. See Updating the compute node configuration and Installing the operating system for details.

If you have changed the configuration of the compute node or if you are installing a different compute node from the one that you removed, you must configure the compute node through the Setup utility, and you might have to install the compute node operating system (see Using the Setup utility).