Re: [opensuse] Failure Trends in a Large Disk Drive Population [ Was: About Backing Up]
At 04:53 PM 2/20/2007 -0500, you wrote:
On 2/20/07, David Brodbeck <gull@gull.us> wrote:
Stevens wrote:
Auto manufacturers try to predict how their interiors and their paints will last, too, but until both are subjected to the Texas sun they are only
Sure. As anyone who's ever had a couple of hard disks fail can attest, MTBF numbers are mostly fiction.
I don't know about that, it is amazing how often disk drives fail shortly after the warranty expires.
Actually, the Google paper on this is very interesting. http://216.239.37.132/papers/disk_failures.pdf
I've rarely had a drive fail on me before I could get the data off it. I had a drive near to failure, yet still was able to order a replacement and transfer the data, before it became unusable. How? By using SmartMonTools http://sourceforge.net/projects/smartmontools/ Since using this, I have had no data loss at all. The old name of the project was SmartMonTools, I should be calling it S.M.A.R.T. now. ATA Drives, both serial and parallel (and most SCSI drives) have been self-monitoring and accumulating their own statistics and performance metrics, for the main reason of predicting failure, for decades. That's right, the drive itself monitors itself. You just need to get the data and schedule tests. But the only place most of us have ever heard of it is in the BIOS of many modern machines, where there's a single question as to whether you want it on or not. All THAT does in most cases is perform the most basic test (drive health check) that S.M.A.R.T. is capable of, and only when you boot. How often do you boot your Linux machines? I though so. On the other hand, if you have S.M.A.R.T. (Self-Monitoring, Analysis and Reporting Technology) the smartd daemon is continuously watching your drives for signs of failure, performing non-invasive tests every (insert your desired test frequency here) and logging any anomalies. I run short self-tests every night, and long ones on the weekend. This tool can "see" the individual drives on RAID cards from a limited group of cards: 3Ware, CCISS, Highpoint. I like being able to get stats on each drive, even though they are a single array. Works on many Firewire, USB and other drives that adhere to ATA/ATAPI-4, -5, -6, -7, and -8. Works on older drives too, but the information it returns is less useful. But that shouldn't be a problem unless your drives are more than a dozen years old, maybe even older than that. I think the ATA-4 spec dates from March of 1996. SmartMonTools runs on: Operating System: Cygwin (MS Windows), 32-bit MS Windows (95/98), 32-bit MS Windows (NT/2000/XP), All 32-bit MS Windows (95/98/NT/2000/XP), All BSD Platforms (FreeBSD/NetBSD/OpenBSD/Apple Mac OS X), All POSIX (Linux/BSD/UNIX-like OSes), FreeBSD, Linux, NetBSD, OS X, Solaris, Win2K, WinXP, IBM OS/2 License: GNU General Public License (GPL) [Huh?] Outside a photographer's studio: OUT TO LUNCH: IF NOT BACK BY FIVE, OUT FOR DINNER ALSO --... ...-- -.. . -. ----. --.- --.- -... tpeters@nospam.mixcom.com (remove "nospam") N9QQB (amateur radio) "HEY YOU" (loud shouting) WEB ADDRESS http//www.mixweb.com/tpeters 43° 7' 17.2" N by 88° 6' 28.9" W, Elevation 815', Grid Square EN53wc WAN/LAN/Telcom Analyst, Tech Writer, MCP, CCNA, Registered Linux User 385531 -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
participants (1)
-
Tom Peters