Nagios Support Forum

The IO error could be a server issue as well. There may be a bug with the software running on the server itself.

I did some searching and found the equivalent to the IO check using VMWares tools to rule out the plugin.
Login to the Nagios server and run the following
Replace xxx.xxx.xxx.xxx with the IP address of the VCenter server and yyy.yyy.yyy.yyy with the IP address of one of the hosts.

In my test, it printer a lot of information when using the verbose option so you may want to pipe it to a file.

The other system with the aborted cpan upgrade, best thing to do it to finish the install and see it it gets fixed.

It does spit out a lot, but I used to work on API development so I recognize it right away. Unfortunately it's giving a "Host YYY.YYY.YYY.YYY not found." and I made sure to pick a host that had VMware tools installed just in case. And yeah Nagios servers have the tools installed also.

I really need to find a way to get this plugin to pull the info cause I've got about 100+ VMs that I need just very basic monitoring on, CPU, MEM, PWR, and DISK, for really simple reasons like; I need to monitor the disks for other teams and increase their sizes as needed.

I would have just added each of the VMs individually, but that would put me into the next tier for licensing, and since I don't need to monitor services, processes, or network bandwidth etc... for any of these ones it seems excessive to goto the next license tier on.

The setting --host yyy.yyy.yyy.yyy in the command is not a guest VM but a host that hosts the guests.
For some reason, I cannot get IO to work anymore. If I specify a Guest, it fails, if I don't specify a guest it works as I think it runs the IO check against the hosts.
I swear it worked on our old 5.5 servers but the 6.x fails now.
I have tried different version of the check_esx3 plugin and even a different plugin and that one failed as well.
I wonder if something changed on the VMWare server software causing it to fail using IO.

It seems that way. When trying to fill in the HOST as yyy.yyy.yyy.yyy I get the same result. The end result of the VMware monitor though will tell me the IO/Disk Usage of each VM though right? Cause if it's just the host disk usage, not the VM, then it's not useful to me.

So I get that this is a third party plugin, but it's one that Nagios has natively added to their XI distributions, so does that mean it's supported and updated by the Nagios developers, or how does that work if this is an ongoing issue?

I don't think the issue is with the plugin, I think the ability to get that info using the plugin / VMWare Perl DSK has been removed from the server.
Even the VMWare Perl SDK fails so that it why I think it is removed.

There is another method for polling the VMWare guests that seems to work but it requires a vMA server to be installed on your network to run the checks.
Take a look at these links for more information.
https://exchange.nagios.org/directory/P ... re/details
https://exchange.nagios.org/directory/A ... rd/details

Ok, so we're not crazy. It just really is not working after 6.0. Thanks for all the help and info tgriep. I'll review these and find the alternative I need. You can go ahead and close this post if you like.

You can go ahead and close this post if you like.

I am locking this thread.