Poor performance with VMware vCenter interface

[khdetw]

We have configured Nagios XI trial instance to interface with our vCenter server as per
http://communities.vmware.com/community ... e_sdk_perl

When our approximately 250 VM's with 5 or 6 monitored services per VM were connected the Perl API consumed immense CPU resources on the monitoring host, with performance grinding to a halt. Apparently the VMware Perl API does not scale well. It creates a separate check_esx3.pl process for each service on each host, so you can do the math on that. Has anyone else encountered this and how was it solved?

I believe the VMware Python API may not have this issue. Is there a VMware implementation for Nagios XI that uses the Python API?

Someone had previously mentioned the box293 approach described at https://exchange.nagios.org/directory/P ... re/details

The box293 plugin looks promising however it relies on the VMware vSphere Management Assistant (vMA). We can not use the vMA due to considerations with an appliance that fails to meet our agency security requirements and patching schedules. Apparently VMware is discontinuing support for vMA, due in part to the same concerns that our security folks have with it. https://blogs.vmware.com/vsphere/2017/0 ... ation.html
I attempted to contact Box293 support about this over a month ago and received no response.

[dwhitfield]

I attempted to contact Box293 support about this over a month ago and received no response.

How did you try to contact him? Did you try @box293?

How often are the checks running? If it's not required to check every 5 minutes, changing the defaults on that would help immensely.

You may also want to check out our general performance tuning document: https://assets.nagios.com/downloads/nag ... ios-XI.pdf