Date: Tue, 24 Jul 2007 11:26:04 -0700 From: Jeremy Chadwick <koitsu@FreeBSD.org> To: Josh Paetzel <josh@tcbug.org> Cc: Bill Swingle <unfurl@dub.net>, freebsd-stable@freebsd.org Subject: Re: problems with Hitachi 1TB SATA drives Message-ID: <20070724182604.GA3759@eos.sc1.parodius.com> In-Reply-To: <200707241230.53119.josh@tcbug.org> References: <46A54B6F.9010100@dub.net> <20070724044208.GA79101@eos.sc1.parodius.com> <200707241518.35730.doconnor@gsoft.com.au> <200707241230.53119.josh@tcbug.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, Jul 24, 2007 at 12:30:49PM -0500, Josh Paetzel wrote: > I don't have any experience with the Hitachi 1TB SATA drives, but I > know an outfit that was trying out the Seagate 1TB drives and had 8 > out of 12 fail their burn-in (a 3 day torture test) My luck with > consumer SATA drives has been incredibly dismal, with ~40 of them in > service I see multiple failures a year, including drives being DOA > and dying after a few weeks of service. I wouldn't be at all > surprised if one or both of the drives was bad right out of the box. > It could be something else of course, but don't discount the fact > that they could be bad from your troubleshooting just because they > are new. This is good advice, and I considered including it in my (long) previous Email. I removed it at the last minute, however, because the only evidence shown in the mail was ad4 behaving oddly. Some off-topic-of-thread facts worth pointing out, as well as some experiences I've had with disks failing out-of-the-box, and a recent failure that cost me quite a lot of data (some of which financial): * If both of those drives were bought at the same time from the same place, chances are both came from the same fab and were manufactured at the same time. Disk fabs have historically been proven to have "batches" of bad disks (if you want sources I can likely dig up articles discussing it, most of which are confirmed by the manus stating they had a manufacturing process that was questionable from X date to Y date). So, there is much higher chance of both of those drives being bad if they were bought at the same time, vs. if you bought each drive separately at different times of the year. * Hard disks are growing in capacity, but are not growing in physical size. We're pushing 1TB in a 3.5" form factor. And the same applies to laptop (2.5") drives. The margin of error continues to increase as we try to cram more and more data in such a small medium. I personally would *love* to see drives go back to using a 5.25" form factor, especially for large capacity disks, since chances are it means higher reliability (read: less chance of error). * There's a lot of common Internet talk about SATA/PATA drives being less reliable than SCSI. My own opinion (based on years of experience with both workstations and servers) is identical. I've run "old" 36GB SCSI drives for years with, at most, 1 grown defect; while in comparison, I have replaced more PATA drives than I can count. SATA drives fall somewhere in-between (less overall failures). Example: Three months ago I bought a new 500GB Seagate SATA disk and had it fail during the initial newfs. Seagate's own tools determined the disk did indeed have bad blocks. I RMA'd it. The refurbished replacement I received (thanks for not sending me a new drive!) died about 3 months later, and I lost almost 300GB of data, and of that about ~100GB was irreplaceable. It's my own fault for not doing backups [see below]. Another RMA. To add fuel to the fire, Seagate *again* sent me a refurbished drive (and I used their advanced replacement program), which has sat in a box unopened since received. I ended up buying two WD 500GB drives to replace the single Seagate; one of the drives is used for nothing other than doing incremental dump(8)s of the other (and the main OS drive). If either of those 500GB drives fail, I'll be able to recover in some way somewhat painlessly. * I'm left questioning why a disk manufacturer would process drives (by this I mean the manufacturing process) differently based on their transport type. It would cost a *huge* amount of money to have separate fabs for SCSI, SAS, and SATA/PATA. It also would make no sense to have employees/workers handling/building SCSI disks "more carefully" than SATA or PATA. I would assume they're all handled in the same way. But I've never worked in a HD fab, so this is speculative. * All this leads me to the topic of backups. Hard disks are growing in capacity at a rate which the backup industry cannot follow. It's getting to the point where you have to buy hard drives to back up the data on other hard drives, but anyone with half a brain knows RAID is not a replacement for backups. There is presently nothing __affordable on the consumer market__ which makes backing up 300GB+ of data easy. Everything that's capable of doing this is in the tens of thousands of US dollars, if not more. Am I going to sit around once a week backing up a terabyte of data to ~120 dual-layer 8.5GB DVDs? Nope. The closest thing out there right now is a product from IOMega called REV, which (at most) offers 70GB of storage per disk, or 140GB with compression. A new IOMega REV (which includes one 70GB disk) costs US$600 MSRP. You read that right. * SCSI is outrageously expensive even in 2007. I have yet to see any shred of justification for why SCSI costs so much *even today*. It costs only a smidgen less than it did 15 years ago. * SCSI is on its way out. Seagate recently announced that they'll no longer be supporting SCSI products, possibly by the end of next year: "Seagate has announced that by next year they will no longer be supporting SCSI product and will be moving customers to the SATA interface." http://www.horizontechnology.com/news/market/market_perspective_storage_04-11-2007.php I'm willing to bet others will follow suit. -- | Jeremy Chadwick jdc at 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?20070724182604.GA3759>