What is the best way to install and configure an Axon EE server for a high workload (handling thousands of commands)?
The default configuration of Axon Server is already capable of handling thousands of commands per second. Typical bottlenecks are the command handling time in the client applications. If commands are injected at a higher pace than the clients can handle them, it may lead to queues filling up in Axon Server. If the AxonServer queues have reached maximum capacity, Axon Server starts rejecting new commands until some pending commands have completed.
The queue size is configurable in Axon Server, but increasing the capacity should not be needed, as it would not solve the underlying issue of command handling being too slow.
On the Axon Framework side, you can check the command bus capacity. If the usage is high, you can increase the number of command threads in Axon Framework by setting the “axon.axonserver.command-threads” property (default value is 10).