Overview
The article outlines steps for upgrading clusters to version 6.0.
Procedure
Preparation Before Upgrade
1. Additional ports need to be opened: 50050, 50051, 50100, 50101, and 50200. Failure to open these ports may result in startup failure, with the BI process disappearing directly.
2. For clusters with the process separation parameter enabled, you need to disable ('DistributedOptimizationConfig.spiderConfig.spider_engine_mode', 'local') by deleting it.
Upgrade
1. Close all project nodes.
2. Execute the upgrade tool on the master node (the node that stores the updated database data) and follow the steps in the document to upgrade to version 6.0.
3. After a single node can be accessed normally, copy the WEB-INF, scripts, and upm folders in the /webroot directory of the node where the upgrade tool was run to other nodes. If the WEB-INF folder is large, only copy the dashboards, lib, plugins, and reportlets subfolders. Then you need to delete the cluster.properties file in the WEB-INF/config folder of each node.
4. Start the other nodes and check whether the cluster status page is displayed correctly (usually, one node is in sync status, and the others are in recover status because the data is being synchronized). If everything is normal, you can start accessing the project.
5. Modify the nginx configuration to implement random request forwarding. However, because version 6.0 already includes random forwarding function, this step can be skipped.
For details, see: Extracting Cluster Nginx Configuration
6. The example file nginx.conf.zip
shows how to simplify the configuration process for version 5.1 cluster and implement random forwarding request. You need to modify the IP address and port number according to your specific environment.
Places that need to be adjusted based on the specific environment:
