Since upgrading my DS211 from DSM 3.0 to 3.1 in the summer of 2011 (now at the current build of 3.2), I've been experiencing random corruption (primarily security descriptors) on the NTFS volumes (formatted via Windows 7) on the iSCSI LUN I have JBOD-configured across a pair of Western Digital Caviar Green SATA drives.
I access the volumes/LUN from a couple Windows 7-based boxes across fairly reliable 802.11n wireless connections. Most of the failures are occurring or at least detected because of the online backup services located on one of the two boxes are reporting failures to run over night.
I've been successful in repairing the corruption so far with chkdsk and in some extreme cases with an icacls-based script I wrote.
I have go so far as to shrink the original volume that had spanned across the LUN and create another NTFS volume since my current storage needs are 20% of what the combined drives provide. Unfortunately the corruption has occurred on this new volume as well.
I'm trying to avoid replacing the drives because of the cost and all physical indicators (SMART, bad sectors or lack of, heat, air flow around the unit and responsive during access) appear to show they not a factor.
Any constructive feedback is appreciated.