Before you deploy OceanBase Cloud Platform (OCP), OceanBase Developer Center (ODC), or OceanBase Migration Service (OMS), you must add the required servers to OceanBase Admin Toolkit (OAT).
Prerequisites
Before you add a server, make sure that you have installed and logged on to OAT as a super administrator or an O&M engineer. For more information, see Install OAT.
Procedure
In the left-side navigation pane, choose Server > Server Management.
In the upper-right corner of the page, click Add Server.
On the Add Server page that appears, enter the following server information:
Basic information, including Server IP, SSH Port, IDC, Credentials, and Description. The following table describes the fields.
Field Description Server IP The public IP address of the server to be added. SSH Port The SSHD port of the server to be added. IDC The IDC where the server resides. You can select Add machine room from the drop-down list to add an IDC. Credentials The access credentials of the server to be added. At present, two types of credentials are supported: username + password and key. Description (optional) The description of the server, which can facilitate server management and maintenance. Server initialization settings.
If you want to take over a component or product, which means a service is already running on the server, you can disable the initialization option to skip the initialization procedure.
A new server must be initialized. Otherwise, you cannot deploy OceanBase services on the server.
The following table describes the initialization fields.
Field Description Server Usage The purpose of the server to be added. OceanBase OS User The OS user that installs and runs OBServer and OceanBase services. A custom user is supported. Password The password of the OS user. UID:GID The UID and GID of the OS user. Docker Root Directory The data directory of Docker. Clock Synchronization Specifies whether to enable clock synchronization. Automatically Synchronize to OCP Specifies whether to synchronize server information to OCP for management. <main id="notice" type='explain'> <h4>Note</h4> <ul> <li> The default OceanBase OS user is admin. A custom user is supported. </li> <li> By default, the password is empty. You can specify a password or click <strong>Random password</strong> to generate a random password. </li> <li>Pay attention to the following points when you specify the OceanBase OS user and its password: <ul> <li>If the OceanBase OS user does not exist and the password is empty, the system creates a user without setting a password. In this case, you cannot log on as this user in SSH mode. Instead, you must log on as the root user and then run the <code>su - username</code> command to switch to the required user. </li> <li>If the OceanBase OS user does not exist and the password is not empty, the system creates a user and sets a password for the user. </li> <li>If the OceanBase OS user exists and the password is empty, it means not to change the password of the OceanBase OS user. </li> <li>If the OceanBase OS user exists and the password is not empty, the system changes the password of the OceanBase OS user. </li> </ul> </ul> </main>
Clock synchronization is disabled by default. If you enable clock synchronization, you must set the clock source IP address, minpoll, maxpoll, maxslewrate, and specify to use this server as a secondary clock source (optional). The following table describes the fields.
Field Description minpoll The minimum interval between time synchronization requests sent from the chrony client to the clock source, in seconds. maxpoll The maximum interval between time synchronization requests sent from the chrony client to the clock source, in seconds. maxslewrate The maximum slew rate for the clock of the chrony client, in parts per million (ppm). Use this server as a secondary clock source After this option is selected, the server can provide services as the clock source before requests from clients on the allowlist are allowed and requests from clients on the blocklist are blocked. This parameter is optional. If you enable Automatically Synchronize to OCP, OAT automatically calls an API of OCP to synchronize server information that you specified during initialization to OCP for future use. By default, this option is disabled. After you enable this option, you must select an installed OCP. If no OCP is available, this option is disabled.
Note
The installed OCPs refer to those available on the page that appears after you choose Product Services > Products > OCP in the OAT console. For more information about how to install OCP, see Install, upgrade, and uninstall OCP.
Disk initialization settings. To ensure the stability of OceanBase Database and other OceanBase services, we recommend that you configure the directories of the services on independent disks or partitions. If you select the disk initialization option, OAT automatically creates the corresponding volume group (VG) by using a logical volume manager (LVM), and mounts the logical volumes (LVs) of the corresponding size to the specified mounting directories.
- Select a disk or partition: The system automatically identifies unmounted disks. You can select one or more disks to mount.
- Mounting Purpose: Select a mounting purpose. Valid values: Software Installation Path, Data Disk Path, and Log Disk Path.
- Mounting point: Select the mounting path based on the actual situation.
- Space (GB): Specify the space size based on the actual situation, in GB.
Click Submit.
If disk initialization is enabled, a dialog box that shows the check items appears after you click Submit. If all check items pass, click OK to add the server.
If disk initialization is disabled, the configurations are directly submitted and the system starts server initialization after you click Submit.
Then you can view the new server in the server list. The server is in the Initializing state. On the Server Management page, you can find the new server and click View Tasks in the Actions column to view the initialization progress. At the top of the page, you can search for a server by IP address.
If an error is returned for a subtask that does not affect business, you can view the logs and manually mark the subtask as success.