Shutting down a node

You can conveniently shut down a host or SCVM in an ACOS cluster via AOC. Shutting down an SCVM in AOC does not affect the power state of the corresponding ESXi node.

When replacing hardware components, such as memory, NICs, CPUs, or storage controllers, put the node into maintenance mode, shut down the node via AOC, and then proceed with the offline maintenance. In the case of an emergency requiring the node shutdown, you can forcibly shut down the node without enabling maintenance mode for it.

In a cluster, you can only shut down one node at a time.

Risk warning

When a file storage cluster is deployed in an ACOS (AVE) cluster, for each file storage cluster, if you shut down two or more nodes where the file controller is located, the file storage cluster and the file storage service it provides might become unavailable.

Prerequisite

• The AOC virtual machine is not running on the current node.

• The current node is in the Healthy, Initializing, or Abnormal state.

• After the current node is shut down, the remaining number of nodes in the cluster meets the requirements for electing the master node.

• The cluster is not using the management IP address of the current node to associate with AOC.

• The node is currently in maintenance mode.

  If the node is not in maintenance mode and is forced to shut down, ensure that the following conditions are met:

  • If there are virtual volumes in the cluster whose redundancy policy is erasure coding, in addition to the current node, the number of nodes in the cluster that are with healthy storage services and are not being removed is not less than the number of data blocks (K) configured for the erasure coding of any virtual volume.
  • There is sufficient free storage space in other nodes in the cluster, that is, the following two conditions are met:
    • The available storage capacity on the remaining nodes in the cluster is more than the used storage capacity on the current node.
    • The available write cache capacity on the remaining nodes in the cluster is more than the used write cache capacity on the current node.
  • The current node has no data with only one replica.

Precaution

• If the node is not configured to enter maintenance mode or if you forcibly shut down the node while it is entering maintenance mode, the virtual machines on the node will be forcibly shut down. If the virtual machine has HA enabled, it will first trigger HA. If the virtual machine has enabled Autostart with host, it will automatically start when the node is started.

• If static data encryption is enabled for the cluster, the cached keys will be lost after the node shuts down, and they will be cached again upon startup.

Procedure

1. Log in to AOC, select the target node in the node list of the cluster, click the ellipsis (...) on the right, and then select Shut down or Shut down SCVM.

2. Enter the reason in the pop-up dialog box, and click Shut down or Shut down SCVM.

Next steps

If you need to shut down or restart the ESXi host, after shutting down the SCVM, follow the prompts in the Shut down SCVM dialog box and the VMware technical documentation to shut down or restart the ESXi host on the VMware page.