June 14, 201115 yr I was wondering if anyone here had played with power-up-in-standby? I'm building a 26 drive machine, so I'd like to use power-up-in-standby so that I don't end up with having to spec a monster PSU... I'm not having a lot of luck with libata controllers and Hitachi drives - if they're behind a Silicon Image port multiplier, they don't get seen (and sometimes all the other drives disappear too). Hitachi drives connected to Silicon Image 3132 or 3124 controllers don't get recognised either. Direct to an AHCI port (Intel H67, or Marvell 88SE9123) seems to be OK. linux-ide thread: http://marc.info/?t=130762548800002&r=1&w=2 Cheers, Tim.
June 14, 201115 yr But you will still have to have all drives running when you check parity, or when you shut down the system as the filing system is closed down. So you still need to be able to safely power all drives running at one time.
June 14, 201115 yr Author Yep, but the drives use around 8 watts when active, but a max of 32 watts when spinning up. Whilst PSUs are built to deliver short-term peak currents in excess of their continuous load currents, I'd guess that asking a PSU to push out 4x the continuous current for the time the drives take to all spin up is probably a bit high, and I'd guess that being able to deliver 2x for short periods would be more usual. Some rough calcs would lead me to believe that power-up-in-standby should mean that I can spec something like a 500w PSU instead of one which is nearer 1kw. Of course, I'd need to do some careful measurements and tests to see if that's actually the case, but power-up-in-standby can't hurt - if it was useless, then the drive manufacturers wouldn't go to the trouble of implementing it... Tim.
June 14, 201115 yr What S80_UK said. You WILL need a PSU what can handle spinning up all drives for a parity check, power down, etc. If you use all green drives that is 2amp per drive so a total of 52 amps just for the drives. Something like the CMPSU-850TXV2 should do for all green drives, plus will give you a little ability to mix and match some 7200RPM drives in there. If you will be adding more 7200RPM drives then green drives, then I suggest getting a bigger PSU. DO NOT scrimp on the PSU, get something that can power ALL your drives and still leave you some headroom. Running 26 drives is going to take a beefy PSU.
June 14, 201115 yr As already said, the very first "read" failure on ANY disk will result in the immediate spin up of ALL disks to re-construct the sector that could not be read. There is no way they will be spun up sequentially and if they were mostly all sleeping, the current draw is significant. I'll put it this way... a marginal power supply causes hair-loss. (You'll be pulling your hair out trying to figure out seemingly unrelated issues.)
June 14, 201115 yr Author As already said, the very first "read" failure on ANY disk will result in the immediate spin up of ALL disks to re-construct the sector that could not be read. There is no way they will be spun up sequentially and if they were mostly all sleeping, the current draw is significant. Hmm, that's a PITA, and does make PUIS pretty much useless. It'd be nice if the kernel dealt with that more gracefully - e.g. by having a mechanism to set the maximum number of drives which are allowed to be brought out of a spun-down state at any the sane time. That'd require a few days of coding tho, which I don't have time for at the moment tho' - not going to get much in the way of savings on a single server either... Actually, on this box I'll have multiple arrays, but that means all disks in a particular array will be spun up (I plan on having multiple arrays with this server), but still, it'll be significant, as you say... I'll put it this way... a marginal power supply causes hair-loss. (You'll be pulling your hair out trying to figure out seemingly unrelated issues.) Yep, I've dealt with such issues in the past, and was planning on doing some supply testing with a current clamp and digital scope. Incidentally, I know that some WD drives have an option to limit the max-spin-up-current, but I can't find any way to do this with Hitachi drives (albeit I did find a Hitachi patent mentioning this feature), unless anyone knows any different? Tim.
June 14, 201115 yr Author S80_UK said: the CMPSU-850TXV2 should do for all green drives. Thanks for the recommendation - I'm using 2TB Hitachi 5K3000s exclusively. The CPU is a 34w TDP Core i3 3100T, so current draw outside of the drives is not going to be much. Tim.
Archived
This topic is now archived and is closed to further replies.