From mboxrd@z Thu Jan 1 00:00:00 1970 From: Bill Davidsen Subject: Re: Sleeping hard drives in an array? Date: Sun, 04 May 2008 20:04:12 -0400 Message-ID: <481E4EFC.2050805@tmr.com> References: <29a863790804300802i358ab6d9t2be907b47176bd5b@mail.gmail.com> <481B836E.6040208@tmr.com> <481B918E.4@panix.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <481B918E.4@panix.com> Sender: linux-raid-owner@vger.kernel.org To: berk walker Cc: Greg Cormier , linux-raid@vger.kernel.org List-Id: linux-raid.ids berk walker wrote: > Bill Davidsen wrote: >> Greg Cormier wrote: >>> Is it possible to sleep hard drives in an array? >>> >>> I have a HTPC at home that's on 24x7. It does all my torrenting as >>> well, and is mainly a media server. >>> >>> Can I sleep the drives in my RAID5 array while it's not being used? >>> It's an XFS partition. >>> >>> I have each drive set as >>> >>> hdparm -S240 /dev/sdX >>> >>> But I'm fairly sure they are not spinning down :( Is there some >>> activity mdadm is doing in the background? >>> >>> These things are probably idle 22 hours of the day. >>> >> >> I have several very similar systems, and I note that on one the disk >> light blinks every five sec or so, while the others don't do that. >> The one that blinks is the only one running LVM, all the others were >> partitioned by hand. Does that apply to your system, LVM in use? >> > Geez!! pardon me if I don't either "get it" or am unable to > communicate about "it". Those thinking that I'm un-necessarily > repeating myself - email your flames and I will be gone from here for > a few months. > > Put / [and whatever ELSE that needs to be mounted] on a USB drive. I > recently got a 2Gb drive for pocket change!! > > Learn, or borrow someone who knows how - scripts. I don't think that > you would give a darn if a ram-drive didn't "spin-down". So PUT what > can not be put to sleep on something that doesn't have mechanical wear!!! > > Your system can be set up to WAKE up with everything mounted and ready > to go on ANY trigger/situation that YOU want - except the states need > to be either on or off... there are lots of states that newer 'puters > monitor. I'm sorry but THIS is not rocket science. > > NOW!! YOU FAILED TO RESPOND!! To my earlier response to you. Please > allow me to ax you again.... WHY do you want spin-down? HMMM a - I didn't, the original poster did b - his question was why it didn't spin down, so the WHY is irrelevant c - none of the comments below are relevant to why his array won't spin down > Less noise- - buy new drives > Less elect. co$t - If they aren't seeking, draw is LOW > Less wear on the drives - I have never had a drive mfg > respond to my questions about this. My [totally > personal] view on this is if you're shutting down for a > few hr., you will lose, weeks will win. Rest of post snipped, also not related to why the array won't spin down. The point about drive wear could be addresses by looking for studies of same, it appears the bearing failures are related to POH, positioning errors to total seeks, and electronic failures to power cycles. The data I have is some years old, but I doubt that's changed, but it's not relevant to the spin down issue. And no one seems to have answered to my query on use of LVM causing frequest access. -- Bill Davidsen "Woe unto the statesman who makes war without a reason that will still be valid when the war is over..." Otto von Bismark