Forgive me if my query can be answered by common knowledge.  After some searching in the forum and wiki, I wasn't able to locate solid enough info to allay my fears.

 

Here's my situtation:

 

I have an UnRAID array currently running 4.6 that has been running admirably for a few months.  It has 7 drives.  1 Parity, 5 data, and 1 unassigned.  The 2 largest drives are 2TB, that would be the parity drive and the largest data drive.  It is currently setup with user shares only.

 

The array froze up yesterday, and when I rebooted, drive 1 (2TB) was marked bad.  I've determined that the drive has failed.  I do not have (and sadly cannot afford) a replacement 2TB unit.  That said, I have more than enough free capacity on the other drives in the array, including a 1.5TB drive that is entirely empty.  Here is my plan of action.

 

-Login to the UnRAID server locally and manually move/merge the contents of drive 1 to the other drives (this is currently in progress/very time consuming)

-Stop the array

-Unassign the failed drive

-Power off and remove the failed drive

-Reboot and use the "initconfig"  (SP?) command to reform the parity data without the bad drive

-Start the array

 

If my assumptions are correct after reading the material I have, this should result in no data loss and a non-degraded (albeit smaller) array when I'm done.  Is this correct?  Are there any other gotchas or issues I should be aware of?  And lastly, having moved data between the drives locally, will all the data still show up correctly when browsing via samba?

 

Thanks in advanced for any insight.

That'll work fine.

 

Peter

There are a few comments at the end of thus thread that should help, as well.

 

http://lime-technology.com/forum/index.php?topic=11861.0

Thanks for the insight, gentlemen.  I'll mark the thread resolved once I've completed the process.

Sorry for the late update.  The operation was successful, but due to the data involved took a week to completely migrated, etc.  Sadly I've moved on to a new issue in another thread, but I wanted to mark this one as solved and once again thank everyone for their help.