Date: Fri, 20 Jun 1997 23:37:46 +0930 (CST) From: Michael Smith <msmith@atrad.adelaide.edu.au> To: jkh@time.cdrom.com (Jordan K. Hubbard) Cc: chat@freebsd.org Subject: Re: Announcement: New DPT RAID Controller Driver Available Message-ID: <199706201407.XAA04999@genesis.atrad.adelaide.edu.au> In-Reply-To: <27545.866803027@time.cdrom.com> from "Jordan K. Hubbard" at "Jun 20, 97 03:37:07 am"
next in thread | previous in thread | raw e-mail | index | archive | help
Jordan K. Hubbard stands accused of saying: > > Erm, I think the difference we're probably seeing with wcarchive is > the fact that each drive bay doesn't hold more than 8 drives (most > holding far fewer) and is independantly bolted to the rack. That > probably provides pretty good vibration isolation, but undoubtedly also > at a drive-to-housing cost ratio which would be unacceptable to the > truly big RAID builder. That depends. The truly big RAID builder would be a total bannana if they didn't have an airconditioning engineer on the team, and there'd be lots of screaming about BTU's per cubic metre if the picture you're painting was as extreme as it is fun to imagine 8) > I would also be suspicious of the cooling properties of a plastic > drive enclosure (it seems like packing it in a mini-Igloo ice chest > would be no worse to me ;-), Again, depending on your definition of "serious", if you're going to depend on the thermal conductivity of your mounting hardware, it's only a small step to a small chiller unit and some water-cooled mountings. This would let you get your drive density back up, and save on rack volume that you would otherwise have dedicated to airflow. > Unless, of course, they had something like the AMES wind tunnel > providing forced airflow past the drives, then I suppose the plastic > sled construction wouldn't really matter much, would it? :-) High airflow velocity doesn't guarantee the best cooling; often this just gets you lots of turbulence near the drives and pushes the bulk of the air right past without actually picking up much heat at all. Oops, sorry, rambling again. Most of this is pickup from reading I did last year on stuff related to keeping a rack full of decidedly hot mixed-signal (low-power RF, logik, high-power RF) boards cool in a sealed room. In the end, they installed the heat exchanger for the A/C unit near the floor anyway, and the whole exercise was wasted (now they just have problems with snow and stuff coming in the window...). Moral of the story? Reinforcement of the "the best solution is rarely obvious, and often different from last time" concept that others have put forwards... > Jordan -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199706201407.XAA04999>