If your VMWare checks experience timeouts or are unusually slow, you may be running out of available entropy (available random numbers) on the XI server. These random numbers are necessary for the VMWare Perl SDK to function properly/quickly. Random number generation is expensive in terms of CPU cost, so you can see delays until there are enough random numbers to fulfill the random number request.
You can validate if you are running out of entropy by running the below command when you are experiencing slowness with the check_vmware_api.pl plugin or any other plugin that uses the VMWare Perl SDK:
NOTE: You may need to run it multiple times to catch the issue but the number should not be close to zero.
If it is close to zero you are likely running out of entropy and you'll need to install a service to fix the system (see below):
NOTE: If you get any errors related to not having a hardware RNG device, you'll need to look at adding a hardware random number generator to the system before rng-tools will work properly:
Feb 10 17:15:04 xxxxx rng-tools: Starting Hardware RNG entropy gatherer daemon: (Hardware RNG device inode not found)
Feb 10 17:15:04 xxxxx rng-tools: /etc/init.d/rng-tools: Cannot find a hardware RNG device to use.
Run the below commands to solve the available entropy issue:
Currently supported versions of the Enterprise Linux (CentOS/RHEL/Oracle):
sudo yum install rng-tools
sudo systemctl enable rngd
sudo systemctl start rngd
sudo apt install rng-tools
sudo systemctl enable rng-tools
sudo systemctl start rng-tools
After running these commands, run the entropy test again to see if the issue is resolved.
Article ID: 916
Created On: Fri, Feb 11, 2022 at 10:31 AM
Last Updated On: Fri, Feb 11, 2022 at 10:41 AM
Authored by: rspielman
Online URL: https://support.nagios.com/kb/article/vmware-checks-timing-out-or-slow-916.html