Suggestion: Disks Option, add Spin-Up delay to reduce peek current

[Ericloewe]

Really? We are not talking of dimensioning a bridge or safety question. So load balancing is not engineering discipline? When an engineer has to build a toy factory he should dimension it regarding just November/December production request even if the rest of the year, demand is 10 times lower? Not average dimensioned and production balancing over months?

Toys can be trivially stored, rectified, regulated and filtered 12V can't.

All engineers that design nowadays laptops are wrong because cooling system is made to be effective on TDP values, while CPUs overlap this value during boost time.

That has nothing to do with the point. Besides, CPUs can throttle to stay within TDP with no serious issues - a PSU can't tell a bunch of hard drives to wait for more power.

You're missing the point. Spinning down drives is a Very Bad Idea (tm) in most cases. If you're worried about your PSU's longevity due to the extra spin up activity, you sure as hell are in one of those cases.

 

[Bidule0hm]

All engineers that design nowadays laptops are wrong because cooling system is made to be effective on TDP values, while CPUs overlap this value during boost time. Intel himself is wrong using inertia of cooling system for this turboboost because during this time it overtakes cooling capabilities?

Yes, because they just want to reduce costs and space usage. But it's more or less ok because it's a desktop PC, not a server designed for reliability.

So even with a huge delay of 5 sec, as you suggest, we are talking of tens seconds.

It's unacceptable for me too. If you need to wait tens of seconds to access a file you'll be annoyed really quickly... and Windows doesn't react well to this kind of things.

You can do a survey if you want and you'll see that a lot of members say the same.

 

[jgreco]

Really? We are not talking of dimensioning a bridge or safety question. So load balancing is not engineering discipline? When an engineer has to build a toy factory he should dimension it regarding just November/December production request even if the rest of the year, demand is 10 times lower? Not average dimensioned and production balancing over months? All engineers that design nowadays laptops are wrong because cooling system is made to be effective on TDP values, while CPUs overlap this value during boost time. Intel himself is wrong using inertia of cooling system for this turboboost because during this time it overtakes cooling capabilities?
Power supply has to start all drives together (boot up) so IT IS properly dimensioned. We are just talking of improving stability and reducing load on system.
No need of suppressing I/Os, no need to create any dependency between drives, just a delay before sending ATA commands. in case of long idle. It already exists something similar called NCQ (but instead of re-ordering, you would just add tempo). And no need to exaggerate, I gave an example, and because your power supply has to start all drives together, you most probably can group your hard drives in 2 or 4 groups. So even with a huge delay of 5 sec, as you suggest, we are talking of tens seconds.
YOU find unacceptable, and your statement, with some other contributors, is that what you find unacceptable is unacceptable for everyone. So whatever.

We actually ARE talking of a safety question; some of us come from a background where if the computing systems we've designed fail unnecessarily, people can actually die as a result. The lesser safety bar here is that if a PSU fails, it could do damage to the remainder of a very expensive storage system. Hard drives are not particularly tolerant of voltage sag.

There's already a comprehensive discussion linked that talks about why properly sizing your PSU is advantageous. The flip side to that is that no one's actually been able to produce an explanation of why deliberately undersizing a PSU just for the sake of doing so is advantageous.

"Just a delay before sending ATA commands" is suppression of I/O pending a condition. When you're creating a system to do something like that, you need to create a system that limits the rate at which drive starts would happen. This could be simple FIFO queueing or something more complicated such as an understanding of the dependencies in the transactions about to occur. This absolutely gets more complicated on a larger system with multiple power supplies, because then you really do have to move to a model which can take into account dependencies on power supplies (such as where you have multiple shelves).

Five seconds is not a huge delay; we've seen drives with spinup times longer than that.

As for comparisons to Intel turbo boost, there is plenty of reason to assume that thermal inertia of a heatsink can absorb a short burst of heat (it's called basic physics), but there is little reason to believe that having a 400W power supply momentarily try to drive a 600W load is not doing damage to something somewhere (it's called basic electronics).

So anyways, yes, we find it unacceptable. But the glorious thing about open source software is that you are free to prove us wrong. Please do feel free to go and create a fork of the project where you can show us that we're all idiots who don't have any clue about any of this.