@skids A number of backup strategies have been suggested in your topic, I will come to my own in a minute, but while backup is about preserving the original image data, without which no recovery is possible, what is the actual proposed recovery strategy and what “crisis” might it need to address.
Arguably the recovery/reconstruction process dictates what and how the data should be backed up?
Issues that may arise
-
The “simple” need to increase capacity to hold more images and related elements.
-
The ongoing deterioration of a drive (of any sort) that is holding image data (i.e. the image, xmp sidecar, DOP, any other image related sidecar file)
-
The desire/“need” to increase the performance of a unit that is holding some/all of the images.
-
The actual complete failure of a unit holding image (and related) data, the “total loss” scenario.
-
The loss of or damage to the unit holding your DAM database and the database(s) of any other editing software you are using.
-
The loss of the unit holding the DxPL database, since this is a PhotoLab forum I expect that to be an issue as well.
If the DxPL database contains nothing that is uniquely important to your image editing then, arguably, no special considerations need to be made for its preservation.
If it doesn’t contain any ‘Projects’ (Win or Mac) or ‘Advanced History’ (Mac) then its recovery can be accomplished by relocating an earlier or even empty copy of the database to a new or replacement drive, and use the images and the DOPs to rebuild the database as and when deemed necessary, simply by discovery if/as and when that happens or by (re-)indexing.
On Windows it is possible to locate the actual database on any drive and always possible to restore a database from a backup on both Windows and Mac
However, the DxPL database can have a “sting” in its tail but first my own backup “plan”!!??
My own “plan” consists of three machines and 2 USB drives.
Each machine has 1 x 8TB HDD + 2 x 6TB drives and attached to one machine (currently) there are also 2 x 8 TB USB3 HDDs.
In addition I have a 5TB, 4 TB and 2TB portable USB3 HDDs and those are (but not rigorously) used to hold “off” site copies, i.e. hidden in the garage or the garden room.
The machines are synchronised regularly but not automatically.
In addition I have a NAS drive configured with 2 JBOD (Just Boring Old Disks is still my favourite translation) units, my “X” and “Y” drives which hold copies of key personal files, a copy of my software library and a 1920 x 1440 JPG copy of all images which is also on the three machines and the 2TB portable drive but not on the 8TB USB 3 drives (space issues).
My collection of images is 3.1TB of 3.4TB of photos and related paraphernalia, i.e. including scanned family photos etc. and appears to contain 482,991 items.
Although I use the DxPL database I cannot claim to be “settled” in my use of that database neither have I embraced the IMatch DAM I purchased a few years ago nor use the ACDSee database given I have used that product on and off since version 2.4!
Earlier his year one of the 8TB drives failed and later one of the 6TB drives suffered extreme health issues.
Recovering so much data from one of the backup machines was tedious across the 1Gb LAN so I actually removed a drive from one of the remaining machines and attached it and the replacement drive to an old machine and rebuilt it that way so that I could run it 24 hours a day!
My principal mechanism for synchronizing disks is Beyond Compare.
My problems are that I have allowed my backup mechanisms to lapse somewhat.
I need to take some time to put them all back in place and get then up to date and maintain a routine to minimise any potential loss.
Now to the potential “sting” in the tail of DxPL.
DxPL does not rely on drive letters to maintain its links to drives, it uses the Drive Serial number/GUID/UUID (or whatever it is called) in preference i.e. for one of my systems the following is the table for all attached drives
and look in this post to see a rather long discussion of the problem and how it can be resolved PL7.1.0 (Win) Issues when changing a drive with a summary of the procedure at PL7.1.0 (Win) Issues when changing a drive - #3 by BHAYT.
Basically if I opened one of my systems and removed the 8TB drive and put it in place of the damaged/lost drive, with the same drive letter, DxPL would ignore the existing database entries for that drive letter, including any existing ‘Projects’ etc., completely and start rediscovering the data afresh from the “new” drive!?
It is probably one of the dumbest things that DxPL could do and other software manufacturers who use a similar strategy actually offer the user an option to adopt the new drive in place of the original drive for which it has many thousands of entries in its database!
The procedure I described in the post, I referenced above, will help if the preservation of DxPL is important to you or other users.
I am writing a program, one of a number that are in progress (!?) and need to be completed to do the adjustment process of the database for the user. But I have run into a problem because I cannot find any code that works to locate the UUID of a disk in PureBasic!?
So on one of my machines this is the ‘Folders’ Table
for any drive to be accepted by DxPL the UUID in the database must match that in the Table in blue above, which they won’t in this case because I have used two machines to get these snapshots!
The program works fine for swapping the network drives but currently the user would need to enter the string for UUID into the program to replace that in the database.
Recent manual tests of the process with PL8 have run into problems so I need to check the procedure to make sure it still works with PL8.
In addition, if DxPL is a critical element of your photo editing then you also need to secure all the elements of the DxPL to ensure continuity, another program I am in the middle of writing.
As for your DAM that also needs to be protected but there you are on your own.
PS:- Having used the backup it is then possible that an additional backup might be needed!
In my case I copied the backup copies to a replacement drive so the total number of drives available was restored but if I had needed to use the backup in an emergency, i.e. couldn’t wait for a replacement drive to be delivered and the 24+ hours to copy the data then I would have needed a new (replacement) backup.