Drobo thinks there's less space than there is.


Dashboard tells me 1 have 1.78TB available, but the Used/Free readings are 989/233 GB. Drobo FW800, 2x1.5TB WD drives and a 500GB drive. 4th slot empty.

When trying to complete a ccc backup I get a ‘low free space’ and ‘drobo can’t protect from driver failure’ warnings. Also had a short relayout Green/Amber flashing lights. Drobo is now back to green but free space is still wrong.

Prior to this I deleted a sparceimage that was incorrectly sized and wanted to recreate it from scratch but deleting the 800GB file only resulted in about 300GB reduction on the Drobo leaving 500GB unaccounted for. Something is screwed up but Disk Utility doesn’t say there’s a problem.


Sounds like a you have a drive misbehaving (hence the relayout).

If you send logs to support they should be able to tell you what’s going on.

OK, will do. The Free space is creeping up very slowly. Now it says 989/321GB

EDIT: gone into relayout mode again. I think the 500GB drive must be failing. Now it’s all green again. Why isn’t Drobo telling me to replace the drive???

Seems you haven’t passed the magic threshold (whatever that may be)… best to grab a diagnostic log and send it in to support when it goes green again.

BOTH 1.5TB drives are having problems apparently causing the relayout to halt. I was advised to replace one of them which I have done - 25hours of rebuild left. Lets hope the other 1.5TB drive doesn’t screw up during this.

The 1.5TB drives were bought together so I will from now on make sure my disk packs don’t contain identical drives to try to reduce likelihood of a failure of a pair simultaneously.

Docchris’s strategy of “exercising” drives outside of the Drobo/RAID/production unit upon arrival to weed out any DOA or near-DOA ones is a good one.

Most times you’ll get drives from varied batches (though usually around the same production date) - especially for retail drives. I got a sealed 4-pack of Caviar Green 2TB retail boxes (brown outer box) and the serial #s of the drives contained within were all over the map. Don’t remember if the batch codes were close though, I think two of them were the same, and the other two were different.

OEMs drives probably stand a better chance of being “closer together.”

I haven’t looked at serials, would be interesting if they are close. Anecdotally I have two drives bought together that failed together, I’ll try to avoid that situation in the future and look up DocChris’s method.

The relayout is taking an age - its on day 2 now and the time remaining bears no relation to how long its actually taking. I guess the bad drive still in there is causing the slowdown.

Likely the case… If it’s feasible, I would disconnect any attached devices to ensure there’s no extra workload being placed on the Drobo from the client (indexing, antivirus, etc).

It suddenly sped up and is now complete! Weird. Now for the other drive and another relayout…

It probably (finally) made it past the “problem” areas. Good to hear things are progressing!!