Our company is seeking for a solution for universal monitoring, it seems Hyperic is the choice. I noticed that Hyperic supports clustered deployment(using an universal monitoring server to monitor other servers). The question is that if Hyperic intrinsically supports server redundancy? I mean, in order to make sure there is always at least one universal centralised server running, we would like to have two of them. So, if one is down, all the collected data from agents will be sent to the backup server automatically. Ideally, if the agents could connect to the two servers at the same time, we can do some load balancing stuff. The two servers might need to talk to each other as well, I guess. But after I went through some documents and threads in the forum, I still have no idea if it is possible. However, I did get an impression that it is unlikely to be done without hacking(as only one server ip is allowed in the configuration file for agents ), right?
The subscription offering from Hyperic does have a high availability option. Currently, it allows for multiple HQ servers to be brought up for the purposes of offloading the sending of agent reports. It does not support failover for the HQ UI or agents, though this is something we are looking into for the 3.x series of the product.
If you have any other requirements for your HA setup, feel free to post them here.