We currently have two servers for Adaxes, both with 12gb of RAM and 4 virtual CPUs running on VMware ESXi. One of our servers is primarily used for scheduled tasks/automation and the other one is primarily used for access by our technicians. There is no load balancer between the servers. Both servers have all Adaxes services running (web, backend, and API calls) on them, supporting a user base of approximately 7000. Both servers are running the latest Adaxes version (2023.2).
We have had an increasing number of server freezes on the technician server after the CPU reaches 100% utilization. We currently have about 60 technicians using the server simultaneously.
I have several questions about optimizing our environment that I don't see covered in the documentation:
- What is the optimal set up for multi-server environments?
- How many concurrent connections per server are recommended?
- In multi-server environments running on VMware, what system specifications are recommended on each virtual server?
- Is there any benefit/downside to having a separated web server from the Adaxes backend servers?