Importing vs indexing in terms of Time Machine Backups

One of the several pro´s and cons that I am evaluating before choosing to index or import databases is the performance and integrity of Time Machine Backups.
In order to have the most reliable and secure backuips by using Time Machine - is it best to import or index? Does it matter at all?

I am not a computer expert, but I would guess that backing up items which are stored in the Documents section (index) are more secure and less prone to backup errors than files stored inside a database (imported).

Neither is more secure or reliable.

Makes no difference to Time Machine in terms of backup performance and integrity. Files of Imported documents are stored within the database using Apple’s filesystem.

Have you tried to backup Devothinj databases with a app called Arq?

No. Given the transfer speed and monthly data limits of my Internet service, online backup wouldn’t be feasible. I use Time Machine backups to swapped USB drives, one of which is swapped every few days to my bank safety deposit box. That’s solid insurance’ I can be back up in a couple of hours.

And I prefer being in full control of my dtata, some of which I don’t want stored “out there” because it is quite sensitive information (I used to lead graduate seminars for government officials, and even some of their names is sensitive information).

1 Like

I use Arq for backup. I’ve never actually tried to restore a DT database from Arq though…

Couple of things — you should read about Arq more and the encryption and key control mechanism and their open source recovery tool before deciding it’s not for you.

Additionally, Arq can use a local store, it doesn’t have to use a storage service provider in the scary cloud. It can just as easily use a local Thunderbolt/USB/Firewire drive or network share on your NAS or anything you can reach via sftp.

It provides far greater accuracy and integrity than TimeMachine, and in the case of my usage rolling backups and snapshots onto ZFS, I never worry about it anymore. It was so trivial to recover individual files with full complete metadata when I needed to, and I also use Arq to back up my research data in DEVONthink to Google Nearline and Amazon S3 in addition to a local store on my NAS at home. It’s amazing software.

1 Like

I’ve rolled back a Database a few times with zero trouble and I haven’t specifically yanked an entire DEVONthink database out before but have recovered a Lightroom catalog a once or twice.

I am in indexer myself, and I use Time Machine for backups.
christopher-mayo.com/?p=2376
christopher-mayo.com/?p=962

Indexing can be a little tricky, but it is worth it for me, because I have greater access to my files (stored in SpiderOak, a cloud service) and I can easily use my files in multiple apps. Arq sounds interesting, and I will take a look at it, but I’ve been satisfied with Time Machine on the rare occasions when I have to track down something that has broken bad. It’s free, exceedingly simple to use, and secure (I encrypt everything).

I can’t say whether importing or indexing is “better” (it depends). It seems to me that it’s best to understand what you will get out of each one in your particular use case. Either way, you’ll probably be fine :slight_smile: