11-08-2017 06:21 PM
I configured Network Advisor virtual managers module to point to our 2 virtual centre servers (each VC server manages 20 ESXi hosts across 4 clusters) last month.
Since doing so the CPU on the VC servers has gone from ~40% with to 2vCPUs to utilising 100% on 4vCPUs, and users experiencing timeout errors in connecting to VC.
Would this situation be known to other implemetors of integrating the two systems, and what are the recommended system modification for integrating NA with VMware vSphere?
I've turned off NA for now so we can resolve the system P2 incident, but wish to be able to turn it on again with the right modifications. to maintain system performance.
Any assistance would be appreciated.
11-14-2017 07:48 AM
By default I think BNA is polling the vcenter every 30 seconds. I am aware of a patch which modify the polling behaviour of BNA to reduce the effect, but you need to contact the support provider for your BNA version, if still applicable for your version.