Originally posted by: Brentium4
All - We have experienced an issue in our shop that has us confused. This issue seems to affect both SI and TC-Sync (that's what we use). When servers with large NTFS volumes are splt and put into PSUS some we've observed occasions where the data is not the same on the disks. Even though the SI or TC pair states pair in the array (which I trust). One fix we've found that works is dropping the NTFS volumes and resyncing.
Our issue has affected our primary file server during a DR test and now management wants answers. I believe the data is on the platters and it is an OS related (the file server runs Windows 2003 - 32bit). The OS seems to have outdated information on the files that rescanning and rebooting will not resolve.
Has anyone seen this issue before? If so what best practices do you use to prevent it. Also can you explain why this might happen? For our DR test our HBAs are in a disabled configuring until after the pair has been split.
Thanks for any insight you can offer into this issue.