I'm curious as to what the actual error messages for the compress were. I can't think why you would get a "database full" message and the only reason for a locked message is if something else was using it. Trying to compress the database while the calibre jobs were still running might cause that.
For finding the problem books, you should be able to do a search for recent files when connected to the device. If you sent 100 books, they would all have roughly the same timestamp. The order they are processed in has changed over time. I think 1.9.17 was processing them in filename order. But, it might have been the full path.
|