Nice. Are your databases that large or do you plan to add a lot of data in the future?
My database is currently 1TB in size, and I will continue to add data. I was recommended Devonthink Pro by a German colleague five years ago. Your software was the only reason I upgraded to A MacBook with M2 Max chip, 96GB RAM, and 4TB SSD. I bought Devonagent PRO as well.LOL
With that setup, I expect you will enjoy your computer a lot, but would you mind updating us over the coming months?
I have a large collection of databases as well (about 2 TB), but I am rarely working with or syncing any individual database larger than a few hundred GBs. My MBPro (Intel i5 / 16GB RAM) is a wonderful computer, but it is getting a bit long in the tooth, and work becomes a bit cumbersome for me (lag) if I don’t keep my “working” database relatively small. At some point in the next year, I expect I will be upgrading, and it would be nice to know what is possible at the extreme end of the spectrum.
I have the exact same setup now and have been using DevonThink Pro since 2015. I use both DT2 and DT3 on two different laptops (I have four DT licenses). I use about 21 different databases (collectively totaling 600 GB) with the biggest one being an archive of 20 years of email, 300 GB. The one thing I’ve learned over the years is to always buy the largest internal hard drive you can. I used to juggle database files from a backup drive to my older laptop with a too small of an internal hard drive. Also, I’m finding when I’m low on hard drive space, say under 20 GB of free space on my main working laptop, my DT3 RSS feeds stop working, though this could be a different problem I’ve yet to solve.
External M.2 SSD enclosures with Thunderbolt 3 interfaces are now available, allowing for fast external access to SSDs. like this
This method has allowed me to achieve over 12T of SSD storage, with speeds that are incredibly fast.
Wow I’m very curious what you use DevonThink for and what kinds of files you are storing that are taking up that kind of space.
I use Devonthink Pro to store my digital library, which includes ebooks, digitized books and documents, notes, literature, web pages, and even backups of some websites. I also use it for my academic work, which generates a lot of file copies and versions. All files eventually end up in Devonthink Pro.
When your main data type is text and the volume increases to the TB level, Spotlight indexing can become a nightmare. You have to make a choice between slowdowns and Mac global search.
This has all been quite helpful. Thanks.
In my case, my digital library consumes the most space, with several TB in total of digitized materials (PDFs and photos of archives), class materials for teaching (PPT, videos, etc.), documents, notes, etc. This seems similar to mrchild’s case.
For the databases, I have whittled it down to about 2TB of stuff I “need” for research and teaching on a daily basis. My problem is that my HD is only 500 GB, so I have two “working” indexed databases on my computer and several “archived” ones on my external drive (databases in one folder indexing everything on the HD). In total, they add up to 2 TB. If it was on my HD, that’d be one thing, but split between an external and internal drive using a “working” and “archive” system (even though just about everything is really working), it’s a bit clunky, to be honest. However, it is one way I have found to deal with this much data.
TSlow searches are not a big deal, because I can set the search going with Find any File (a DEVONthink product), DT, Hazel, or Spotlight) while working on other stuff, if necessary. Usually, though, searches only take a few seconds to turn up what I want.
With my next computer, I might bump up the internal storage to take care of this artificial working / archive divide I’ve got. That would be a huge help in my daily workflow, but going up to something like 4TB would cost nearly as much as an iPad. Yikes. I will have to give it a think.
imho RAM is the best bang for the buck
and I can add external storage
Agreed, especially as DEVONthink loves having available memory to work with. The more, the better!
update! -_-
How many items/words do your databases actually contain? This has to be a huge amount of data.
This exceeds of course the recommendations (up to 300 million words per database, up to 300 thousand items per database) “slightly”
Really unusual is the number of unique words. Does the database contain lots of poorly OCRed scans? Or material in many different languages?
And I’d ask…
- Why are you building this monolithic, singular database instead of splitting things into smaller and more focused databases?
- Are you just do indiscriminate data dumps into the database, like whole hard drives, home directories, etc.? If so, that is inadvisable and something we’ve discussed more than once.
yes ,this is a sad story!