Linux system configuration cluster

This document describes how to configure a cluster on a Linux system.

2.1 Basic Environment Preparation

Before deploying the cluster, make sure that you have prepared the environment required for the cluster project: Environment Preparation

Environmental testing: environmental testing tools

Note: WebLogic single-computer clusters and WebLogic dual-computer clusters are WebLogic's own clusters, not Fanruan clustering solutions, and are not recommended.

2.2 Deploying FineBI Projects

In each cluster node server, deploy the FineBI project to the container.

The deployment containers are described in the following table.

2.3 Configuring external database

Select the project on a node as the master node project and configure the external database.

The types and versions supported by the external database are shown in the following table.

2.4 Installing and configuring cluster components

2.4.1 Installing the configuration file server and starting

When a file server is not enabled, inter-node synchronization is used by default to ensure file consistency in the cluster, although there is a possibility of synchronization delays between nodes when there are network communication problems between them.

To avoid such problems, it is recommended to use a file server, which currently supports the following three.

1）vsftpd：Linux system installation configuration FTP

2）sftp：Linux system configuration using SFTP

3）HDFS：HDFS Resource Repository

Note: 「Automatic synchronization between nodes」 is not applicable to multiple nodes, otherwise it will affect the use due to the communication problem between nodes, and it is only used when there are two nodes, and 「File Server」 is required when there are more than two nodes.

2.4.2 Installing and Configuring the Stateful Server and Starting it Up

Redis Cluster can ensure high availability of the stateful server compared to the standalone mode, but it requires certain server resources and increases operation and maintenance costs, so please choose your own.

1）Redis Standalone：Linux System Installation and Configuration Redis Standalone

2）Redis cluster：Configuring Redis clusters on linux systems

2.4.3 Install and configure load balancing and start

Load balancing is an indispensable part of the cluster solution, and is independent of the project. If you have the ability of operation and maintenance, you can choose the appropriate hardware and software load balancing, currently there are customers using load balancing including Nginx, treafik, F5, AWS ELB, NetScaler, Fanruan's basic solution based on Nginx, for non-Nignx load balancing, you must operate their own configuration, refer to the document:Load Balancing Configuration Guide

Nginx：Linux system installation and configuration of Nginx

2.5 Platform Configuration Cluster

Master node project reference Platform configuration cluster Documentation to configure the cluster, when the first node appears, copy the project package (webroot) of the first node under other servers, then start these servers, and the new node will be added to the node management.

Note: For cluster projects registered with local machine information authentication, you need to re-register the cluster project after adding nodes.

3.1 Changing the FTP path

Problem Description

After the cluster project is built, I need to change the FTP file server path due to lack of disk space.

Solution

1) Copy all the files under the original path to the new directory and assign permissions to the new folder. For details, please refer to: Installing and Configuring FTP on Linux System, Configuring FTP Service on Windows System

2) Modify the path of the file server on the platform and save it.

3) Reboot each node of the cluster. See Section 3.7 of the Configuring the Cluster documentation for notes on rebooting the cluster project.