Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

I have a RAID 1 mirror on a Lacie 2 Big USB3. When I break the mirror by pulling out one drive, keep making changes on the remaining drive and finally put the first one back in, the drives stay happily out of sync. The unit doesn't seem to detect the mirror was broken. The manual only explains the steps to take to replace a failed drive with a new drive. Does it only re-mirror if the replacement drive is new? What does it look at to determine that?

share|improve this question
    
How do you know it's staying out of sync? Often RAIDs will just resynchronize in the background, and if the differences are not huge, it may not take very long to copy them. This is different than a "rebuild" which implies a complete copy of the data to the new/replacement drive. –  Ƭᴇcʜιᴇ007 May 26 '12 at 19:11
    
According to the manual the leds give status information which in this case is: "All Normal". If I don't do anything further the unit goes to sleep so it clearly thinks it has no work to do. However, when I look at the drives one at a time I see the changes I made above are only on one of them. –  vhauf May 26 '12 at 19:33

2 Answers 2

Taken from Lacie’s support FAQ:

Do the swap with power on so the embedded OS knows what’s up.

share|improve this answer
    
This is an excellent first answer. It addresses the question asked, provides a reference, and includes enough information if the reference URL goes away. –  ChrisInEdmonton Feb 5 at 19:47

I don't know this drive specifically, but most consumer-model RAID enclosures use very simple logic to determine the health of the array. Usually this is just some metadata stored in the slack space of the disks themselves.

The simple logic only triggers a RAID fault in the following conditions:

  • Two disks are detected but only one of them is responding.
  • There is a read/write error on one of the disks during operation.
  • A disk becomes disconnected or is hot-plugged while the enclosure is powered on.
  • The RAID signatures on each drive differ (one disk will be treated as foreign and will not be used).

That's it. Under "normal" conditions (i.e. you use it the way you're supposed to), replacing a bad disk would trigger a rebuild because the new disk would not have the RAID signature on it.

In your case, since you removed one of the drives while the unit is powered off, none of these fault events gets triggered. When you power it back on with only one drive, the RAID metadata gets ignored (the unit operates in single drive mode). When you put the second drive back in and power it on, the enclosure sees the metadata on both disks is identical and assumes that everything is fine. There is no hashing or checksumming that takes place to determine if the disks are in sync the way a real RAID controller would.

Lucky for you, these "dumb" RAID enclosures also do not do read balancing the way a real RAID controller would either. Data is always read from one disk in the array and the other is basically a live backup, so you're not totally corrupt when you power it on. Make no mistake, though, your mirror is broken and the disk you used on another computer is screwed up. What would happen if you attempted to fail over to it is anyone's guess. The disk should be considered totally corrupt at that point even if another computer says it's fine.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.