With the increase in business data volume and user visits, the originally deployed cluster may not meet the requirements and requires more nodes for higher performance and scalability.
This article details how to rapidly add nodes for a cluster project of containerized deployment through FineOps.
The visualized capacity expansion operation only applies to containerized cluster projects.
It does not take effect for containerized standalone projects or non-containerized projects.
1. Ensure that the project type is Cluster, not Standalone.
2. Ensure that the deployment method is FineOps Deployment, not Non-FineOps Deployment.
Adding a project node requires deploying relevant components on the server through the image in the FineOps image repository. Therefore, the admin shall confirm that the required image exists in the repository.
1. Log in to FineOps as the admin, choose Maintenance Center > Image Management, and check whether the required image exists.
2. Check the image repository address and ensure that the server can access the extranet cloud warehouse if the required image is not present in the repository.
Log in to FineOps as the admin, and choose Maintenance Center > Image Management > Setting to check the image repository address.
Ensure that the server where the image repository is located has access to the following extranet address for automatically pulling the latest component images of the project from the cloud during project deployment.
Domain name: fineops-registry.cn-hangzhou.cr.aliyuncs.com
Port number: 443
3. (Optional) Obtain FineDataLink-related images.
FineOps does not provide image packages of FineDataLink and Kafka by default.
Premium users can contact FanRuan technical support personnel to obtain the FinetKey tool that contains FineDataLink and Kafka image packages and update FineOps by referring to Intranet-Based FineOps Upgrade.
To contact our technical support, you can send an email to international@fanruan.com.
Prepare the server in advance for installing the components required for the new node.
The number of servers to be prepared shall be the same as that of new project nodes.
Each FR/BI/FDL project node should monopolize a server.
No other projects or products shall be installed on these servers if there are sufficient resources.
The project and FineOps should not be deployed on the same server.
Recommended: Ubuntu 22
CentOS 7.3 to 7.9
Ubuntu 18.04.4 and later releases
Red Hat 7.6 and later releases
Remaining free disk capacity: more than 50 GB
Remaining free physical memory: more than 16 GB
CPU: 8 cores or more
Confirm that the default port for the application and component to be deployed is available on the server.
Confirm that the port to be used is connectable and not blocked by the firewall or the security group.
Supported
If a port is already in use, use other free ports.
Not supported
Ensure that the port is free and available.
All important files of containerized projects should be mounted outside the container. Create a mounting directory in the free disks of the main application server A and the component server B in advance.
1. Check the server disk space.
Use the df-h command to find a mounting point with large free space.
In this example, the most appropriate mounting directory for the server is the /home directory.
2. Create a folder.
Use the mkdir command to create a folder. (E.g., create a folder named fanruan in the /home directory.)
mkdir/home/fanruan
Use the chmod command to set the permissions on the folder. (E.g., assign 777 permissions.)
chmod 777/home/fanruan
Since a server can be installed with only one project node, the admin shall add a server for the O&M project as the main application node.
Note: This section illustrates how to add one node to a cluster project. To add multiple nodes, repeat the operations in this chapter.
1. Log in to FineOps as the admin, click the xx icon of the project whose capacity is to be expanded, and click Add Node.
2. Enter the node server information, which is described in the following table.
Select Main Application
The server requires a minimum of 8 cores and 16 GB with more than 50 GB of disk remaining..
Enter the host IP address (intranet IP address) of the node.
Note: You cannot add a host repetitively to the same project.
Enter the node installation directory on the server, which is the mounting directory set in the Setting the Project Mounting Directory section.
The default directory is ~/data, which is the /data directory in the home directory of the server being used.
Note: You can use the user account to access the server in the terminal, and enter the echo $HOME command to view the home directory.
Optional.
If the server does not support intranet access unless the intranet IP address is mapped to an extranet IP address, you can fill in a connectable extranet IP address.
The following operations will be automatically performed after the node is successfully added to the project.
Load the new node into the Docker environment.
Deploy the ops_agent component on the new node.
Monitor the new node.
The admin shall deploy the FanRuan application on the successfully added node.
1. Log in to FineOps as the admin, select the project whose capacity needs expanding, and choose Maintenance > Component Management.
2. Click Expand Capacity on the card of the fr/bi 6/fdl-type component.
3. Select the node set in the Adding the Main Application Node section.
The selectable option is the main application node on which no FanRuan application is installed.
The node with the FanRuan application installed is in grey and cannot be selected.
4. Set the resource limit on CPU and memory for FanRuan applications.
It is recommended that the CPU is not less than 8 cores.
The recommended memory value is the machine memory value minus 8 GB, with an upper limit of 32 GB.
5. (Optional) Modify the port number mentioned in the Confirming the Server Port Availability section.
6. Click Start Expansion.
7. Check whether the new node appears in the component list of the cluster project. If so, the capacity expansion is successful.
滑鼠選中內容，快速回饋問題
滑鼠選中存在疑惑的內容，即可快速回饋問題，我們將會跟進處理。
不再提示
10s後關閉
Submitted successfully
Network busy