Greetings,

We are setting up a 100TB storage cluster, and are looking into various ways of doing this. Two key things we need is OS distribution from the central controller into the storage blades and systems monitoring.

So far Kickstart looks like the best option for the OS management for Ubuntu/Debian and SuSE has AutoYAST.

For immediate system monitoring, we can use Gkrellm. For logging and web-accessible interface Munin is looking really good.

As far as the rest of the system, we'll be testing iSCSI and AoE for performance/stability over the next few months. AoE seems wonderfully simple, while iSCSI has more features (which we will most likely not use).

RAID will be done on hardware level.

Any ideas or suggestions?

Thanks!
Andrey.