Could you be more specific?
I'm not changing any paths or the data physical location from an appliance standpoint.
I'm just moving the contents of $tideway$var/localdisk/ to a separate drive, then remounting that drive in the same place. The tideway services are shut down while I'm doing this.
If I have to open a support ticket I'll do it tomorrow though this would seem to be a fairly straightforward configuration change if anything other than a simple data move.
I got it working.
Here's what I did for those who are interested.
Case: Ran out of space on the root filesystem of a test appliance with a single drive
I installed a new drive and formatted it as ext3 with a temporary mountpoint /mnt/tmp
chown -R tideway:tideway /mnt/tmp
shut down the tideway services
mv data/ snapshots/ /mnt/tmp
once the move is complete:
edit fstab add mountpoint /usr/tideway/var/localdisk
mount the drive again(mount -a)
reboot(not really necessary but I had to fix the network as well so I did)
the tideway services should start normally.
Please note: I did this on a copied version of my appliance in case anything went wrong. You should do the same
That is not the correct way of resolving the problem. Should you ever have to restore your system from a snapshot you will likely end up with an unusable system.
As I said I strongly advise you and others to contact support.
ADDM is an appliance not a general purpose server. If you try and change it's configuration in ways that it does not expect you will either cause it to fail straight away or more problematically at some point in the future. I don't want that to happen so please consider following my advice.
I'll do that next time, the important thing for now is that the data is usable. I've started a consolidation run to another server just in case the server should bite the dust in the future. It saved me a lot of work
If there is a correct way for moving the database if there is no more diskspace it would be a welcome article in the documenation, along with the article for adding a second harddrive.
The procedure is in the docs here.
My concern is for other side effects that you may have caused to your system.
I would advise others not to follow your procedure. It is actively dangerous to use the berkely db commands directly and not through the datastore system and doing so can cause very severe unrecoverable data corruption.