Hello, after working out some small issues I've managed to get a couple instances up and running with Ganeti. Had a few questions/issues though. First, I'm trying to do a live failover without the instance going down. I've followed the howto section about this but it insists on wanting to shut down the instance when failing over. I've read the howto over and over, and even created a new instance to try with. At the first step where it says do gnt-instance failover inst1.example.com, it says a reboot is required. I am running a newer Ganeti version than described on the howto. Would that cause an issue? Another question I had was more general. First off let me say this is my first time starting to work with clusters. I've done a lot of searching and haven't found much documentation about Ganeti. How is load balanced among the cluster? I read something about job scheduler/queuing, but haven't found how to actually manage that. I've tried 'time make' on the debian linux-sources-2.6.18 on both the node it's hosted on and the instance. It was about 10 seconds difference, so pretty much the same. I guess I had assumed the load among the cluster would somehow magically be balanced across the cluster. Anyone care to explain how this works? I appreciate everyones time. Thanks, -nux
Anyone have any information on this subject? I'm trying to determine if this is the software to use for a cluster that I'm setting up. Thanks
