https://bugs.kde.org/show_bug.cgi?id=389848
--- Comment #161 from Porkepix <porke...@gmail.com> --- (In reply to Nate Graham from comment #160) > The LMDB developers have responded and are asking for an affected database > file. > > For anybody who can reproduce this issue, I have three questions: > 1. Is it reliable? If you restart Baloo, does it immediately crash again? > 2. Is your database file below, say, 100 Mb? `balooctl status` will tell you. > 3. Are you willing to upload your database to a public location so the LMDB > developers can debug it? It's located at ~/.local/share/baloo/index For me, the crashes were kinda random, but (I'm not sure anymore), could happen during high usage computer, causing micro-freezes, or intensive writes (might be related, due to swap), especially, I sometimes suspected it would kind of dislike files that were only partially written and there constantly "reevaluated". Only bare feeling as I have no knowledge on how does it work here. If this might be related, the disk is a SSD and is fully encrypted through LUKS, which might cause read and write to be much more CPU-intensive. The file currently weight 2.4G. I disabled baloo quite a long time ago now, as you can see: .rw-r--r-- 2.4G clement 22 Feb 0:33 /home/clement/.local/share/baloo/index It would be hard to upload because, well, I would need to control content for privacy reasons to sanitize data, but it both seems it's a huge amount, and that I have no knowledge of data structure. But I can provide more information if I'm told how to do so. Looking at what you said above, the size is a lot bigger than it should be, so, maybe if the bug was considered to fixed and it was a database corruption the solution would be to wipe that file out? -- You are receiving this mail because: You are watching all bug changes.
