Scenario 1: None of the Controller Enclosures in the Cluster Are Powered On

If none of the controller enclosures in the cluster are powered on, you can use the BMC system's external management IP address to log in to the BMC system and run the power-on command to power on all controller enclosures in the cluster. The implementation process is as follows:

1. Use the IP address of the management network port to log in to the BMC system.

   The way to log in to the BMC system is the same as the way to log in to the CLI of the storage system through the management network port. If the storage system is not powered on, you will be directly logged in to the BMC system after making an attempt to log in to the CLI of the storage system through the management network port.

2. Run the power-on command to power on all controller enclosures in the cluster.

Scenario 2: Some Controller Enclosures in the Cluster Are Not Powered On

If some controller enclosures in the cluster are not powered on, you can log in to the storage system using the IP address of the management network port and run the power-on command to power on the controller enclosures that are not powered on in the cluster. The implementation process is as follows:

1. Use the IP address of the management network port to log in to the CLI of the storage system.
2. Run the power-on command to power on the controller enclosure that is not powered on in the cluster.

Networking Rules

• If a storage system has only one controller enclosure, ensure that:
  • The power cables of the controller enclosure are correctly connected.
  • The maintenance terminal is connected to any management network port on the controller enclosure.
• If a storage system has multiple controller enclosures, ensure that:
  • The power cables of all controller enclosures are correctly connected.
  • The maintenance terminal is connected to any management network port on any controller enclosure.
  • The scale-out cables of all controller enclosures are correctly connected. The maintenance network ports on all controller enclosures are correctly connected according to the standard networking mode.

Networking Diagrams

The cable connections of a single controller enclosure are simple. This section only describes the networking mode in the multi-controller-enclosure scenario.

In the following figures, purple cables are management network cables, and orange and blue cables are scale-out cables. To use the remote power-on function, connect network cables (green cables in the following figures) to maintenance ports on controller enclosures.

If green cables are not connected, after the power-on command is executed, only the controller enclosure where the currently logged-in controller resides can be powered on. Other controller enclosures cannot be powered on.

Figure 5-9 OceanStor Dorado 3000 V6 four-controller networking diagram

Figure 5-10 OceanStor Dorado 5000 V6 and Dorado 6000 V6 four-controller networking diagram

Figure 5-11 OceanStor Dorado 8000 V6 and Dorado 18000 V6 eight-controller networking diagram

Prerequisites

• Cables have been connected according to the standard networking mode.
• The power cables of all controller enclosures in the cluster have been connected.
• None of the controller enclosures in the cluster are powered on.
• This operation should not apply if some controllers in a controller enclosure are not powered on.

Precautions

• Some tools, such as PuTTY, cannot be used to access the BMC system before the storage system is powered on. It is recommended that you use the Git Bash tool.

  Download Git Bash installation package on the Git website.

Procedure

1. Log in to the BMC system on the Git Bash tool using the management IP address as the BMC administrator.
   • You cannot log in to the storage system using an IPv6 address to run a command to remotely power on the storage system.
   • This step is the same as the way to log in to the CLI through the IP address of the management network port. If the storage system is not powered on, you will be directly logged in to the BMC system after making an attempt to log in to the CLI of the storage system through the management network port.
   • The user name and password of the BMC administrator are the same as those of the super administrator admin of the storage system.
   • Assume that you remotely log in to the BMC system of the storage system to be powered on via SSH. If the system displays a message indicating that the login fails due to the invalid ESDSA host key, see How Do I Update the Public Key When a Linux Host Fails to Remotely Log In to the Storage System's BMC System via SSH Due to the Invalid Public Key?

   $ ssh admin@XXX.XXX.XXX.XXX
   The authenticity of host 'XXX.XXX.XXX.XXX (XXX.XXX.XXX.XXX)' can't be established.
   ECDSA key fingerprint is SHA256:RV2Qwdr0wL2XXXXXXXXXXXXXXXXXXXXX.
   Are you sure you want to continue connecting (yes/no)? yes
   Warning: Permanently added 'XXX.XXX.XXX.XXX' (ECDSA) to the list of known hosts.
   
   Authorized users only. All activities may be monitored and reported.
   admin@8.46.161.136's password:

2. Run the cliset -d clusterpoweron 1 command to power on the storage system.

   iBMC:/->cliset -d clusterpoweron 1
   Cluster power on OK.

Prerequisites

• Cables have been connected according to the standard networking mode.
• The power cables of all controller enclosures in the cluster are connected.
• Some controller enclosures in the cluster are not powered on.
• In the single-controller-enclosure scenario, if a controller in the controller enclosure is powered off, remote power-on cannot be performed.

Procedure

1. Log in to the CLI of the storage system using the management IP address of any controller enclosure that has been powered on in the cluster.

   You cannot log in to the storage system using an IPv6 address to run a command to remotely power on the storage system.

2. Run the change user_mode current_mode user_mode=developer command to enter the developer mode and then the minisystem command to enter the minisystem.
3. Run the eam.sh poweroncluster command to power on the rest controller enclosures in the cluster.

   Storage: minisystem>eam.sh poweroncluster
   admin:/diagnose>eam poweroncluster
   Power on the cluster successfully.