cnMaestro On Premise - Capacity / Resource Usage

We have been having some issues with our cnMaestro On Premise instance being a bit of a resource hog and operating quite sluggishly even after adding addtional resources. It also has been behaving oddly, losing communication with devices that are actually up and operating with zero issues.

So my questions are:

1. How many devices can one instance realistically support? We currenty have 3,619 devices onboarded, a mix of PMP 450, ePMP, cnPilot R and E series.

2. Does a high availability setup offer some load balancing properties?

3. Has anyone else run into similar issues?

Hi

 For  device number upto 4K, we suggest to have 4 vCPUs and 8 GB RAM. In case of proxy thru controller is enabled, suggest to double up the resources. 

cnMaestro On-Premises supports Layer 2 High Availability through an Active-Standby (1+1) architecture 

Please refer the "cnMaestro On-Premises 2.3.0 User Guide.pdf" which can be downloaded from support site 

https://support.cambiumnetworks.com/files/cnmaestro/   for more information on hardware resources and HA 

1 Like

Thank you for the response. I have seen the user manual but HA does not necessarily state (clearly) that it distributes the load between the two instances in a normal monitoring state (outside of config/firmware pushes), is it purely a redundancy measure?

Is it better to turn up another cnMaestro On Premise server and split the devices? This would make deployment, management however very difficult.

Hi commorat24 -- it is 1+1 Active/Standby, so one system handles the device management, and the second is in standby mode. cnPilot Enterprise devices tend to use more resources than Fixed Wireless devices, and RADIUS Proxy also adds additional overheads. In the past we have seen additional issues when very old processors are used: they are able to support the virtual machine infrastructure, but their benchmarks are very poor.

Hi, i’m having issues, I have 8K epmp, what do you recommend?