Hi,
I will try to answer your questions with as much details as possible.
1. Our communication protocol is optimized to the best way of our knowledge already. We have put a lot of effort into minimizing the amount of traffic that a gateway generates. Of course it depends on the amount of devices and monitors and how often tests are performed, how much traffic that is actually generated but I can assure you we have done what we can to keep it to a minimum.
2. In our lab environment, we have a few gateways with different types of links and one of them is partcularly slow over a VPN (300-400 kbps). We don't currently have information saved on how much data being sent and received over a period of time, but the amount is small, that much we know. Each gateway will display how much data has been sent/received since the last restart of the server so with that information, it's possible to calculate approximates.
3. Any WAN optimizer would not understand the protocol since we are using password authenticated SSH2 sessions.
4. I'm not sure exactly what you mean here. Do you mean for the OS device templates, or for all templates? We have tried to make them as generic as possible out of the box so they can be applied to most versions. The new device templates will be adding only monitors that it can find. The template includes a number of monitors possible to add, then when it's applied, it checks if the value that the monitor supposed to retrieve is possible to read. If it is, the monitor is added and if no value could be read, the monitor is simply just not added to your device.
5. It might be that the eventlog monitor in the Windows internal device template is associated with Server 2003, but definitely subject to change. I will look into that and make sure it's updated for the next release. Thanks for letting us know.
6. CIM is not a method for retrieving performance counter, it's mainly health indicators where you can find out the health of a fan, a power supply, a disk or a memory slot for example. Datastore utilization is a monitor we are working on and hoping to include in the next published build. At the moment, we have had a SSH script that can monitor datastores but it requires SSH to be open on the ESXi host. I would recommend waiting for our datastore utilization monitor that we will actually be testing this week.
7. If you have a look at our presentation from Las Vegas (www.ustream.tv/.../32171731) maybe you can tell me if those are attractive to you ;) In that case, I can login to that server and do some screenshots for you.
Regards,