Date: Sun, 16 May 2010 14:01:51 -0700 From: Jeremy Chadwick <freebsd@jdc.parodius.com> To: Simun Mikecin <numisemis@yahoo.com> Cc: freebsd-fs@freebsd.org Subject: Re: Quick ZFS mirroring question for non-mirrored pool Message-ID: <20100516210151.GA82487@icarus.home.lan> In-Reply-To: <657328.88413.qm@web112409.mail.gq1.yahoo.com> References: <4BEF2F9C.7080409@netscape.net> <4BEF3137.4080203@netscape.net> <20100516001351.GA50879@icarus.home.lan> <alpine.GSO.2.01.1005151937300.12887@freddy.simplesystems.org> <4BEF4A73.8060905@netscape.net> <20100516015850.GA55302@icarus.home.lan> <657328.88413.qm@web112409.mail.gq1.yahoo.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, May 16, 2010 at 01:37:57PM -0700, Simun Mikecin wrote: > ----- Original Message ---- > > From: Jeremy Chadwick <freebsd@jdc.parodius.com> > > To: Kaya Saman <SamanKaya@netscape.net> > > Cc: freebsd-fs@freebsd.org > > Sent: Sun, May 16, 2010 3:58:50 AM > > Subject: Re: Quick ZFS mirroring question for non-mirrored pool > > > > Also, please reconsider going with Western Digital > > RE4 2TB drives. These drives are all "GP" (Green Power) drives, which you do > > not want. There have been numerous reports on the FreeBSD mailing lists > > about problems with these drives (repeated head offloading/parking > > causing problems in RAID arrays), and yes, it applies to Enterprise class > > drive as well; WD has indirectly confirmed the problem in one user's case > > by sending him a "fixed" firmware. I can point you to threads if you > > want to read them. > > > Well, there are two types of WD RE4 enterprise drives: "RE4" and "RE4 > GP". Only "RE 4GP" is a "Green Power" drive. "RE4" is a performance > drive spinning at 7200rpm (and I believe it is currently fastest > enterprise 2TB drive on the market). You're absolutely right. I reviewed WDC's site before making my statement, but the way they do their layout is a bit confusing. "RE4" appears in the 3rd column, while "RE4-GP" appears in its own column, since they segregate it due to its lower power usage: http://www.wdc.com/en/products/index.asp?cat=2 The price difference between these drives is US$50 (RE4 costing more): WDC RE4: http://www.wdc.com/en/products/Products.asp?DriveID=732 WDC CBk: http://www.wdc.com/en/products/Products.asp?DriveID=733 What do people get in exchange for paying 1/5th more in cost? Basically nothing: * TLER is usable / adjustable on both RE4 and Black drives. It defaults to disabled on the Black. Some tech forum weirdos state WD disabled the ability to adjust the feature on later Black drives; I've personally confirmed that's not true (said weirdos probably forgot to disable AHCI when using the DOS utility to adjust TLER). This is probably the most useful feature for these drives when used in RAID or a RAID-like (ZFS) fashion. * RAFF sounds like marketing schmooze, given that both the RE4 and Black drives already provide an extra-stable motor shaft ("StableTrac"). Yes, I've seen the Youtube video of an administrator showing the effects of vibration on array performance/stability; I've a hard time believing this would matter in the OP's system. * PMR has been in use since 2005 on all types of drives, across multiple vendors. There's nothing amazing about it at this point. * More marketing: the Caviar Black lists "dual processors". Whatever. * As has been proven time and time again, MTBF means jack squat since it's all hypothetical (mathematically calculated based on fab tests). Drives will fail no matter what; that's the entire reason people are using ZFS to begin with. ;-) The only feature that sounds even remotely useful is the "dynamic fly height" feature of the RE4, and it's not worth US$50. I have to wonder if marketing these days has found a way to pray on obsessive-compulsive disorder or what... -- | Jeremy Chadwick jdc@parodius.com | | Parodius Networking http://www.parodius.com/ | | UNIX Systems Administrator Mountain View, CA, USA | | Making life hard for others since 1977. PGP: 4BD6C0CB |
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20100516210151.GA82487>