Tony wrote:
Fordem’s response is very valid here, but on a slightly different note, you may get some relief by spinning the drives down. Keep in mind that spinning disks up is very stressful on the hardware and will greatly increase the possibility of a hardware failure. I’d not recommend it, but if it’s important to you, maybe you can give it a try.
Most modern BIOS implementations support disk spin down on some level. Not all will be compatible with NASLite running, but most should do. The trick is to set the drive spin down in the associated BIOS and prevent NASLite from hitting the drives during status updates.
In other words, disable the SMART monitoring for the local drives in question or set the status updates event to 8 hours or so.
One can spend some time testing settings for optimal results, but the goal is achievable on some level by using the BIOS facilities.
I know that this is not your ultimate goal, but it is an option. Hope that helps.
Folks, I really appreciate the quick response and the efforts you've expended on this

But my primary question is still unresolved.
Perhaps better illustrated:
1. Assume a "master server" is built and is running 24/7.
2. Now assume that I've found a way to power up or shutdown slave servers at will (might be pretty stinky...which is why I won't reveal methods here. Might give the designers a heart attack).
Will the master server still recognize the slave servers and accept their "exported shares" even when they are switched on and off the network? I've read everything I can find (including the manual) - but this issue remains a mystery. Just what *is* the "discovery process" for remotely hosted storage? How robust is it? How long would it take to recognize that "server x" is now online and available?