Do the indexed files still exist, what’s their file type and size?
Today I made another check, the result is still (almost) the same, and here are my observations:
-
There used to be 17 files without checksum, all PDFs, but now there are only 16 PDFs left, one new error is an email I added today with drag-drop into the sorter; there is no checksum yet.
-
Three files with differing checksome are part of my Zotero storage folder, the PDFs are not large (0.2KB, 1.5MB, 2MB). The files exist. There are no problems, and I would like to have a command that says “repair checksum”…
-
The other 13 PDFs are also all indexed and somehow made it into my database; although I try to keep it free from indexed files. (It turned out that one email is missing with a 14.5MB attachment)
3.1 Size: 1.3 MB – 76.9 MB
3.2 Existent: none at the location mentioned in the log
3.3 Existing in Zotero elsewhere as Duplicate (recognized by Devonthink): 7
3.4 No Duplicates mentioned but existent in Zotero (and in DT) elsewhere, without being recognized as duplicates: 6
3.5. All PDFs sit in a dropbox folder which cannot be connected any more, I am just getting rid of my dropbox this year. But this seems not to be the problem
So, the problem seems to boil down to the following:
- Three files should be OK, are existent, not too big, but throw a problem.
- 1 file has not yet a checksum, it seems to take several hours or days.
- 13 PDFs are indexed and not connected, so a problem should arise. But I cannot imagine why this should be the problem of checksum.
- I cannot check easily whether the indexed files are connected or not, neither in the info panel nor in the log. I have to check the path manually (do I oversee something?). If the problem is the fact that the data are not connected, that should be mentioned in the log, at least in my opinion…
So this is the situation. I can get rid of all missing entries in DT now, do you need any more infos about them?
Thank you very much.
Thank you for the info! Is the path of the indexed items which still have no checksum actually still valid?
After opening a database DEVONthink tries automatically to calculate eventually missing checksums in the background (e.g. as the app or the computer might have crashed or as the database was created/edited by another version not yet supporting checksums).
In the end if the paths are valid and DEVONthink is allowed to access the files, then there should be actually no more missing checksums soon after opening the database.
OK, i restarted the computer, this morning checksums of all mentioned files (17) were different, so it seems that the path was OK now and the new checksum calculated upon opening.
Then I deleted all entries outside of the Zotero storage folder, because I could get the original files etc., and did not need the entries any more. Then I ran check and repair and then integrity check.
Three index entries (3) from the Zotero storage folder still give a wrong checksum. The PDFs are there. What can I do now?
Does File > Update Indexed Items fix this? By the way, did you just index these files on their own or their complete enclosing folders?
No, I have updated the indexed items several times now, it does not help.
They are all three part of the Zotero storage folder (in my user folder, not synched with iCloud like Documents or Desktop), which is indexed and updated from time to time. I have no clue why these PDFs are different.
A rebuild should definitely fix this but might require some time depending on the size of the database. Other options are to modify the document (in DEVONthink or externally) or to reindex the file(s).
OK, I just marked a word in each of the PDFs, and the next check for integrity went well. Thanks for the advice.
Recommendations for others:
When encountering such a mass of checksum errors and checking each entry in the log file, it might be best to just change something in those files that do exist. When running the check for integrity once again, only those with lost files will be left, and can be treated with more ease.
Request for the log file:
A log window with more options to sort and filter would be very much appreciated, I have suggested that earler, but it may be too much work. So, another, maybe less demanding request: Can we have an option to select entries in the log file and then with one click, give them a tag? One could then go through those files easily from the tag in Devonthink.
Big databases:
Yes, mine is, beside the Mail Archives, the only one, it covers all, which is a good thing. But it is big, and new synchs, rebuilds etc. cost a lot of time, a real drawback, but having one database is worth it I think.