Hi wizards. I''m trying to explore the limits of lustre failover in real deployments. Basically, I need to know what to write down for our customers, which means I need to know what kinds of configurations to test. If you''re using lustre failover, or even have tried it, can you tell me what you did to make it all hang together? In particular, the kinds of things I''m interested are: What HA-type software are you using on the hosts? LinuxHA? Slurm? Something else? What drivers and hbas are you using? Did you do anything special about driver and/or hba configuration? (I''ve been getting a crash course in all the stuff you can configure in that world!) What controllers and disk drives are you using? Again, what, if anything, did you do by way of configuration on the controllers to make them behave? What management software? Ie, when events happen, what software hears about it first, and how does it get handled from there? What configurations or other things, in all of the above, did you try and had problems with? For starters, we expect all our storage systems to interface via fiberchannel, but I''m interested in hearing about all. Thanks in advance for any info.