Devon3 Memory Hog

I paid a very steep $ amount to upgrade to DT3 with the promise of better processing usage than DT2 and I am sad, as well as very frustrated, with the utilization of memory. Everything activates the usage of memory. I am super disappointed. It does not matter how good your software is, and it is good, if I can’t use it! I am also considering asking for a refund until you get this version in order.

The memory usage of version 3 should be more or less identical to the one of version 2. How many items/words do your databases contain, see File > Database Properties? And how much memory does it actually use?

As to memory use, it goes up to: 1.48GB and freezes. As to the number of items: I have 10 databases. the largest 3 I cannot tell you the number of items because opening each individually will cause the memory level to spike and freeze so that I cannot access File.

What is the filesize of the databases in the Finder (though this isn’t a perfect comparision)?
Are you importing or indexing files?

I have 13 databases with a zillion of indexed files on each and the only slow-down is when DT3 is launched and an operation is donde for first time. “See Also & Classify” takes some seconds first time you select it, but once selected changing documents takes one second or less. Same with right click.

The only “slow” feature is search, but, well, searching over 13 databases with thousands of documents into them could take 10/20 seconds, that I think is fast.

I have a MacBook Pro 2017 16GB RAM, 512SSD, and an iMac 2017 24GB RAM 2TB Fusion (both SSD) disks.

Perhaps your problem, @shiiko is your Mac has a mechanical disk.

The freeze makes me think it’s a bad chip that doesn’t get engaged under normal use…

You could check your hardware via integrated test hardware tool:

Some MacBook Pro shows error in one address (only one) because a bug in the tester that is not aware of some devices mapped in memory (like my old MBP Retina 2012). If you get a memory error in a range, your Mac has hardware problems.

The number of items is actually relatively small. Are these huge documents? How many words do these databases contain? Or do you use fuzzy searching?

I’m not complaining about search speed. For me is fast enough. And no, I’m not using fuzzy search (at least not normally). For your information,

Model: iMac18,3
RAM: 25,8 GB, Processors: 8
Startup Volume: APFS - 1,4 TB of 2,1 TB free
Installation Path: /Applications/DEVONthink
Inbox: 1815 unique words, 5825 total words (max. 1494, pdf), 15 items, 7 MB (max. 3,4 MB, pdf)
Almacén: 1589 unique words, 7339 total words (max. 5370, plist), 414 items, 11 MB (max. 2,5 MB, jpg)
Charles Dickens: 760037 unique words, 34318241 total words (max. 676958, pdf), 120 items, 6,8 GB (max. 307 MB, pdf)
Correo: 409302 unique words, 58834829 total words (max. 2667438, eml), 28645 items, 5,6 GB (max. 41,4 MB, eml)
Feeds: 45368 unique words, 567425 total words (max. 24051, html), 1006 items, 14,7 MB (max. 1,9 MB, pdf)
Help: 4601 unique words, 91274 total words (max. 3392, html), 423 items, 12,1 MB (max. 611,8 KB, png)
Investigación: 181792 unique words, 11246042 total words (max. 2977524, txt), 155 items, 206,5 MB (max. 39 MB, txt)
IyC - Sciam: 4089245 unique words, 274475536 total words (max. 633484, pdf), 5862 items, 70,4 GB (max. 82,2 MB, pdf)
Jules Verne: 742266 unique words, 22886430 total words (max. 521577, pdf), 619 items, 39,5 GB (max. 640,2 MB, pdf)
Libros y Manuales: 710144 unique words, 41220719 total words (max. 839519, pdf), 842 items, 11,1 GB (max. 253 MB, pdf)
Pulps en Español: 573020 unique words, 45129475 total words (max. 458277, pdf), 1894 items, 4,6 GB (max. 63,3 MB, pdf)
Revistas: 1170814 unique words, 83953124 total words (max. 110386, pdf), 1888 items, 121,2 GB (max. 199,2 MB, pdf)
RFOG: 57711 unique words, 776165 total words (max. 63815, pdf), 1004 items, 1,6 GB (max. 568,2 MB, mp4)
Science Fiction: 3788313 unique words, 345852295 total words (max. 540258, pdf), 4843 items, 81,7 GB (max. 381,4 MB, pdf)
Scrapbook: 204844 unique words, 7489066 total words (max. 561914, pdf), 3883 items, 7,2 GB (max. 121,2 MB, html)

Doing some test I’ve discovered that the slow in search is caused when start typing and short unfinished words. If I paste the full search, or disable incremental, search does not takes more than 2 or 3 seconds!! :heart_eyes::smiling_face_with_three_hearts:

Are these databases open all the time or are you closing unneeded ones (I hope so) ?

Some of the databases are quite large and/or contains huge documents (probably books). Is searching faster after performing File > Verify & Repair… on all databases first?

This suggestion might sound a little bit weird but this command accesses & caches all required data of databases, therefore afterwards everything should be as fast as possible.

Only email one… :sweat_smile:

Cris, I’m not complaining about the speed. It is waaaaaay faster than I expect. Once I’ve resolved my duplication files problem I’m in love with the program.

Who had speed problems was @shiiko. I answered him that perhaps could be a HW problem, as his program stops working at very low memory usage (about 1.4 GB). My installation go into 12 GB o higher of memory usage with swap without any complain (by my part and by DT itself).

1 Like

Late to this conversation. For comparison I thought it would be useful to share that I have had the opposite experience–DT3 seems far superior with memory management than DT2. I work regularly with several large databases. I often have open 3 dbs (20 gb, 6 gb, 5 gb [finder sizes]). The largest is predominantly pdf files; the other two a fair mix of text and pdf. Additionally I frequently have 20-40 individual documents open. In DT2 this would consume between 3-4 gb of memory, but was highly variable; sometimes it would spike wildly. In DT3 I’ve seen it hover more between 2 and 3; the only significant spike was on the very first launch, which troubled me enough that I’ve been watching it like a hawk since, and it’s been rock solid.

Running High Sierra on:

Model Name: MacBook Air
Model Identifier: MacBookAir7,2
Processor Name: Intel Core i5
Processor Speed: 1.8 GHz
Number of Processors: 1
Total Number of Cores: 2
L2 Cache (per Core): 256 KB
L3 Cache: 3 MB
Memory: 8 GB



Sorry that I have not replied but was hit by the weekend and a DT database full of duplicates that were consuming my cpu. There should be a better way to deduplicate files and find what are missing files, especially among duplicates. Should have the ability to see missing files within a SmartGroup and not having to go out to a log. This experience was really hair pulling for me.

Model Name: MacBook Pro
Model Identifier: MacBookPro11,3
Processor Name: Quad-Core Intel Core i7
Processor Speed: 2.3 GHz
Number of Processors: 1
Total Number of Cores: 4
L2 Cache (per Core): 256 KB
L3 Cache: 6 MB
Hyper-Threading Technology: Enabled
Memory: 16 GB

Shiiko, don’t know if this might help you - it did help me when dealing with enormous numbers of duplicates (which were my fault). What I did was exporting databases to folders and files, and using TidyUp to find and delete duplicates. I am sure there are other good apps. Then reimporting all to Devonthink, and happy with that.


I understand that this can be frustrating. But before lashing out too hard against the software, please consider that if these issues were generally applicable, this would be a huge topic on these forums. You have to assume that you’re not the one with the heaviest DBs running on a middle-of-the-road machine. Hence, it seems likely there is a local issue with your setup, as others above have pointed out.

I’m running a handful of always-open DBs, reasonably large ones, with a lot of documents with huge word count, and I do that on the lowly 2016 MacBook (1.2 GHz Core m5, 8 GB RAM). Yet things are really smooth on the DT3 side, whereas other software can bring this machine down, as one would expect. I sometimes run nearly a month between reboots, with DT3 open 100% of the time. No memory leaks or the like observed. In other words it’s unlikely that DT3 is the primary culprit in your case.

1 Like

I have an older machine, have DT3 always open, and have noticed zero performance differences. I haven’t looked at memory usage because, absent performance problems, there is no reason to worry about such matters. This is an 11" Macbook Air from early 2014, with a 1.4Ghz iCore5 and 8GB RAM. Nothing special and probably the least powerful mac you could buy five years ago.

I’m extremely pleased with the upgrade to DT3, especially the UI changes, and have never considered performance an issue in the slightest.

1 Like

@rfog How did you create the inserted list of database information?

That’s the information from when you use our bug reporting mechanism…
Hold Option and select Help > Report Bug. You can see the info there. Just don’t send the email unless you’re actually opening a support ticket. :slight_smile: