Hi
Last month, my hard disk (Mac) started having trouble, and the repair shop moved all my files onto Dropbox. Anxiously, I copied my DT databases back into the database folder on my new hard drive and crossed my fingers while opening my main DT database.
I was able to open the DT database. Running ‘Verification and repair’ left this message: “Found 0 inconsistencies, 0 invalid filenames, 60 missing and 0 orphaned files”. Is there a way to see the names of the files that are missing? I have all files stored outside DT, too, so I can import the missing ones again. I have hundreds of files in the database, so finding the missing ones will not be an easy task.
Are you indexing or importing into your databases?
Also, to answer your initial question in the title: "Harddrive failure - how to prepare better for this situation?, don’t you have current local backups?
The repair shop moved the databases xxx.dtbase2 to a Dropbox folder.
You should never put your DEVONthink databases in Dropbox or any cloud-synced location or you could irreparably damage them. This is documented, well-known, and often discussed.
My backups are unfortunately not as frequent as they should be And always to Dropbox.
As for dealing with the database issue, see “REPAIRING A DEFECTIVE DATABASE” section in the DEVONthink Manual, page 193 of the 3.9.7 version for some guidance for dealing with this. No need to repeat all that here.
Until you sort this out, I think you should turn off any syncing so that the defect(s) is(are) not synched to other devices. Synching is not backups.
Also, you may wish to re-think using DropBox synch folders as backup. Any flaw there gets synched quickly to all copies including to Dropbox servers. Backup strategies covered here numerous times along with a short bit of advice on page 19 of the DEVONthink Manual.
Thank you - I will have a close look. And make a sustainable backup plan. As I had to reinstall DT, the sync hasn’t been set up yet. And I now understand that I should wait until things are sorted out here at my main computer.
The problem for me with user manuals is that they feel overwhelming. And quite often, I do not really know what I am looking for.
I understand that I should not sync DT databases through Dropbox.
When exporting a database xxx.dtBase2.zip, I put that on Dropbox. I will not unpack it in Dropbox (if needed), just keep it and hopefully never use it. If I have to use it, I will transfer the zip-file to my computer before extracting the database. Is that wrong, too?
When exporting a database xxx.dtBase2.zip, I put that on Dropbox. I will not unpack it in Dropbox (if needed), just keep it and hopefully never use it. If I have to use it, I will transfer the zip-file to my computer before extracting the database. Is that wrong, too?
Nope. That’s perfectly fine as the ZIP encapsulates and insulates its contents.
However, I hope that is not your only backup. Search our blog for backups.
But my personal data is not located on the Mac’s disk, instead I use a Thunderbolt drive for all my data.
This drive get’s mirrored with “rsync” hourly to a second Thunderbolt drive, similar to TimeMachine.
In addition, the most important directories get mirrored daily to one of two alternating SSD drives (Samsung T5 and T7) to have some sort of archive - I exchange them weekly. Once every 1 to 3 months, I exchange one of those with another T7 that I store in another place.
Also weekly, I upload those important directories to OneDrive, some unencrypted, some encrypted with Cryptomator.
Those important directories include those that I index with DTTG, which in addition get’s synced by DT to two different webDAV locations.
I feel safe
But currently, I have no automated way to backup the DT databases themself. I only do this manually … any idea how to automate this?
Hmm.
Yes, that is true.
But I want a fresh archive of those, in case I break something.
Gettings single things back from Timemachine can be tricky - and there would be no copy on other disks and locations … like for all other stuff I have.
I would love to have such archives automatically and daily
There is a DT script “Daily Backup Archive” but that does not seem to run automatically.
Yessss, I should consult the manual
I have now repaired my database as much as possible, and it is once again a running system.
It turns out that all the files causing issues were of the ‘.rtfd’ type. However, not all ‘rich text documents with attachments’ were defective.
Fortunately, I have all the original files stored on my computer, so I can rebuild my DT database. That said, I wonder if these types of files should be avoided in the future, or was I just unlucky?
Furthermore, I’ve now included an external backup in my weekly routine.
Which is good, but not good enough imo – you’ll still lose changes between those snapshots. I’d rather go for daily (unless you only work once a week with your databases, that is).
LIke @chrillek says, “not good enough for me”. I suggest you partition part of the external disk (if that’s what you mean by “external”) to take a TimeMachine backup that runs automatically. My TimeMachine backups are set to run “hourly”.
Also look to setup an automated backup regime of entire system, which includes the DEVONthink files, using the 3-2-1 regime. Lots of internet articles about that so no need to repeat here.
And, occasionally check that backups running, and do test restores just to make sure that they exist and that you know how to restore.
And always to Dropbox.
