can anybody answer ? please ?
Sorry for the late reply, but just came across your question, hopefully this is still helpful...
We use PRTG to monitor every aspect of our VMware deployments and the monitoring is critical to keeping the server up and running. If you have critical services that you are providing and can't afford downtime then I suggest you make the time investoment to setup PRTG to track your environment. By hosting the solution you can't monitor much of the backend infrastructure, but hopefully your provider is already doing so and may even give you access to their monitoring for free (especially if they use PRTG like we do)! You can see if a hard drive (predictive) failure is imminent so it can be replaced before an outage takes place, you can monitor scratch drive space on the ESXi host in case you increase RAM and the swap disk is too small, you can especially monitor CPU and RAM utilization/growth over time so you can predict when you will need additional resources. You can look at your network traffic utilization by port or across "conversations" using Netflow. Basically the tool has many uses and will become a critical part of your ability to provide high uptimes. You will find that for the money, PRTG is hard to beat and covers about 90% of what you need.
Regretfully though it still seems to be missing some important but basic items, like monitoring the ESXi host for accurate time. If your ESXi host doesn't have accurate time there are many things that can break in your environment, such as Windows Active Directory synchronization and authentication, which makes services and SQL access fail. Even though it is a small item it has big implications and apparently PRTG does not have an answer for this yet (according to their support folks the SNTP sensor does not work on ESXi hosts). They seem to have these strange deficiencies in their product coverage but there aren't that many and for the money, it's impossible to beat...