NSX Manager High CPU

[Stephan_Zur]

High CPU is seen on Manager that has the virtual IP assigned. 

Checking API processes status and monitor processes via CLI on affected manager shows uproxy consuming a lot of CPU. I do not find a lot information online regarding this process. What is this process used for?

NSX Version 4.2.0.0.0.24105817. No Proxy configured on NSX Manager. 

   {
      "process_id": "73774",
      "user": "uproxy",
      "nice_value": "0",
      "virtual_memory_size": 2042.46484375,
      "resident_memory_size": 605.140625,
      "shared_memory_size": 0.0078125,
      "cpu_usage": 359.9,
      "memory_usage": 2.5208079420843577,
      "command": "java"
    }

[RolandS]

That's the Apache Tomcat "proxy" process for authentication and session management. Isn't there too much logins API requests against your manager? 

[Sriram ChunchankatteMelukote]

Apart from the tomcat web server processes, please also check the processes related to the corfu db. Sometimes this can also cause spike in cpu usage on the nsx managers.

[Stephan_Zur]

Thanks for your replies and hints.

It is only uproxy service which is high on Manager node the cluster VIP points to. Corfu db is normal on all nodes. 

Monitoring NSX audit log to identify where API calls are coming from.