Confused on the correct way to backup all my devonthink data, advice appreciate :)

You are setting the same variable twice in this code. I guess that you might want to perform the act of archiving first before changing the variable’s name.

Edit: I also suspect that there’s a typo in the second set: It probably should be “/Volumes…” with a leading slash.

After having a look at the script itself: If you want to save to two different locations, you might want to refactor the part starting from the set timeout line until the end of the save proper into its own subroutine (that might be called a handler in AppleScript, but I’m not sure). Pass in the target name (this_archive) as a parameter and call it twice: once after the first set this_archive, once after the second.

thank you @chrillek Yes, the variable is set twice because I would like to make the same backup to two locations. The second set works without the “/”. Don’t know what is to “refactor” but yes, having a timeout between them seems logical. I did not understand anything of what you mean from here: " until the end of the save proper into its own subroutine (that might be called a handler in AppleScript, but I’m not sure). Pass in the target name (this_archive ) as a parameter and call it twice: once after the first set this_archive , once after the second."

I have recently subscribed to Arq Premium. I’ve set it up to only back up from a folder, where I intend to place all items I want to backup.

Am I ok to simply store my DEVONthink databases in that folder or do I need to do something else?

That should be sufficient. But you should also have a Timemachine or other backup strategy in place to safe all your data.

I already ready run a CCC to external disk. This is essentially for my offsite backup. I was just wondering if DT database worked ok in a folder backed up by Arq.

I’m assuming that I can open and work from the DT database from the Arq folder?

Yes. But best not while the backup is running. I set mine to push the databases to Arq at 2 in the morning, when I’m certainly not at my computer.

3 Likes

Hopefully that’s not a critical requirement
My TimeMachine and Arq backups are running hourly (default)

I don’t know. I’m coming from Unix, and there we shut down all unnecessary activity before running a backup. So, me doing the Arq backup only when I’m not working might just be a stupid old habit. Or not.

Let me put it differently: File systems tend to cache and reorder writes. So, backing up something while you’re working on it could lead to a backup that is slightly different from your local state: You have changed something, but it’s not yet committed to disk when the backup runs. This might cause a problem when restoring.

Or the backup tools are simply clever enough to flush the disk cache before copying the data.

Can you clarify this? It sounds like you mean to work from within your backups.

No, I want my main databases to be in the folder that Arq backs up. So i would place all my main databases there and work from that folder.

I know @eboehnisch is using Arq for his Mac backups. However, I’d schedule the backup to happen at a time when you’re not using DEVONthink or DEVONthink To Go and syncing.

Thank for that. So Arq should only to be used when DT is shut down.

An application that uses a scheduled snapshot-style backup, i.e., it gathers info from one point in time, “freezing it”, then doing its backup in the background should allow for simultaneous use. However, as we always advocate caution on behalf of peoples’ data, the safer option is to do your backups during a period of inactivity.

Technically, this can also free up resources for the backup to happen.

2 Likes

Thanks for that. Is there are better way to backup to Arq?

I’m not using it here, so @eboehnisch or some other Arq user would have to weigh in on more specifics.

1 Like

As Jim mentioned, that’s not really necessary.

I just commented on this in another thread. Arq and Carbon Copy Cloner documentation also states that it doesn’t matter what you do while the backup is running. I even deleted a test database once while a backup was running and was able to restore it.

1 Like

I have Back Blaze and use Time machine backups onto some hard drives. However I keep my data files on a thumb-drive, 64 GB one that is solid state. I think I have used it once, to date I have never had to get anything from Back Blaze. However once woulld be enough to justify the cost. I still really use the 3-2-1 strategy. Three backups of my vital stuff in whatever form, including the thumb drive in my case, Two different ‘types’ of proper backup, not synching systems and the like one must note. And one ‘off site’.
I have to admit the thumbdrive is what I think of as my ‘real’ back up and the simplest and easiest to use. I have nothing else on my computer that is of any value to me other than what is in DEVONthink 3 by the way. So I don’t have things scattered around: even so I back up the whole system in various ways.

1 Like

Any chance to have a kind of incremental backup of a database?

I mean: I’d like to have the exact mirroring of the database on a cloud, with the folders and the files as I see them inside the DT3 App, not as they are stored in the HD.

I may export the database on a weekly or monthly base, but each time I’d have to re-sync the ALL the files, and it becomes cumbersome because I have many Gbs inside.

(hope it makes sense)

So effectively what you are looking for would be an incremental export; that could presumably be implemented by offering an “overwrite/duplicate/do nothing” option on export (currently export will duplicate files already available in the export destination; to my knowledge the user cannot currently influence that behaviour).

Agreed: files that have not changed don’t need to sync again.
Only the new files and the files that have been changed.