Check Event Log for disk I/O subsystem related messages. If spindle is dying it will eventually be reported to abort / delay commands. Use some free tool (Bing-Fu there are tons of them) to get SMART status for underlying spindle. Run chkdsk to see is
there any volume metadata corruption.
Microsoft is conducting an online survey to understand your opinion of the Technet Web site. If you choose to participate, the online survey will be presented to you when you leave the Technet Web site.