In the IP Fabric managed by Apstra, the customer reported a continuous high CPU in some of the fabric devices

The current implementation of Apstra is to use SSH and CLI to collect various telemetry parameters related to the manage fabric devices

The high frequency of those telemetry collections might cause a high CPU on the managed nodes

the CPU utilization might vary from node to node based on the scale and the amount of CPU time required to collect those telemetry parameters

We observed CPU utilization improvement on those devices
After the customer modified the frequency of telemetry collection by changing the interval timers in the Apstra controller