[XPsp2, SATA WD 250gig split into 10gig c:, 20gig d:, the rest e:, Epox 8rda3+ (nVidia2 chipset), athlon 2500, 1gig ram] My main machine spent some time last night complaining of "The device, \Device\Harddisk0\D, has a bad block." while defragging (Diskeeper 8). Diskeeper reported nothing out of order, so presumably any failures were below its realm of expertise. Messages courtesy Event Viewer, source Disk, message IO_ERR_BAD_BLOCK. Running a "chkdsk /r" over it just now (choosing run on next boot) shows zero bad sectors. Has the drive succesfully mapped them out of use, or is chkdsk a poor reporting tool? It's particularly worrying since D is for Documents, and has a large quantity of stuff I've actually created on it. I'd much rather that bits and bytes weren't being quietly corrupted and incorporated into the rolling backups... (which do exist, O cynical reader!) On top of which, I've just found that C: has 24k of bad sectors, and E: has 16k. 4k allocation units. This worries me too. Can/should I RMA? Or is a reformat enough to shuffle the bad sectors out of the way? The disk is somewhere approaching a year old, I think. Cheers - Jaimie
I'd rather be safe than sorry. Download drive diagnostics for your drive; for Western Digital they're called "Data Lifeguard Diagnostics" and are available from: http://support.wdc.com/download I personally quite like Seagate's SeaTools, which has a "Generic Diagnostic" that works with pretty much any drive you throw at it. Available from: http://www.seagate.com/support/seatools/ If you have confirmed bad sectors, then backup immediately and if the drive is under warranty, get it RMA'd ASAP. Just as a side note, I've known drives to last for years with a constant amount of bad sectors (had 2Mb bad on a 1.2Gb Western Dig that didn't get any worse over time), but I've equally known drives where the number of bad sectors increases with every format. If it's under warranty, play safe and RMA it. HTH, Dave
The quick scan showed and "fixed" errors, the long scan will be going for a second hour yet. I suspect I'll have to do a chkdsk /r on each volume afterwards, to rescan the relocated bad blocks. I'll throw that at it next. Thanks for the links. I wasn't sure of the RMA-ability of a tiny percentage of bad sectors, what with the way bad pixels on LCD screens are considered "legitimate errors" by the mfr's. Thanks for the confirmation. Me too, but I've had only sudden-drive-deaths for the last decade or so and I'm out of practice! Thanks - Jaimie