I got an email from Drobo support yesterday with the following instructions:
“Since Drobo1 is only a back up, then instead of just reformatting, you might try starting the array from scratch by performing a Clear Disk per the attached instructions. Once this completes, verify that “Tools” tab indicates that this Drobo is running v1.3.1.
Set up Vice Versa to run on your Drobo and if you hit any problems then run CHKDSK to see if it finds any problems. After that, try generating a fresh diagnostic file. If you get the “I am unable to save the diagnostic information.” message again, then try rebooting the Drobo and the computer and trying again.”
The “Clear Disk” instructions were pretty straightforward: shutdown the Drobo, straighten out a paper clip, pull the USB cable, pull the power cable, stick the paper clip in a little hole in the back of the Drobo while plugging in the power cord, hold the paper clip down until the Drobo is all booted.
I did all that, formatted the virtual drive with the Drobo Dashboard, and wrote about 600 GB of data.
So far, so good.
In an unrelated Drobo event, I had a hard disk fail on Drobo2. I swapped in a new one. The rebuild (of about 2 TB of data), took more than 24 hours. Not on a par with a conventional RAID, but adequate for the intended use.