I am running into this for past few months. Indexing goes fine until Baloo index database grows close to 4 GiB and then it "truncates" it and all the search and file content indexing data are gone. Before: balooctl status Baloo File Indexer is running Indexer state: Idle Indexed 107651 / 556631 files Current size of index is 3,95 GiB After: balooctl status Baloo File Indexer is running Indexer state: Idle Indexed 107651 / 556631 files Current size of index is 29,06 MiB Baloo_file_extractor: ls -l /proc/4215/fd celkem 0 lr-x------ 1 rrr rrr 64 5. úno 19.02 0 -> 'pipe:[25292]' l-wx------ 1 rrr rrr 64 5. úno 19.02 1 -> 'pipe:[25293]' lrwx------ 1 rrr rrr 64 5. úno 19.02 10 -> /home/rrr/.local/share/baloo/index lrwx------ 1 rrr rrr 64 5. úno 19.02 11 -> /home/rrr/.local/share/baloo/index lr-x------ 1 rrr rrr 64 5. úno 19.02 12 -> anon_inode:inotify lr-x------ 1 rrr rrr 64 5. úno 19.02 13 -> /proc/4215/mounts lrwx------ 1 rrr rrr 64 5. úno 19.02 14 -> 'socket:[24037]' lrwx------ 1 rrr rrr 64 5. úno 19.02 15 -> /home/rrr/.local/share/baloo/index-lock lrwx------ 1 rrr rrr 64 5. úno 19.02 16 -> /home/rrr/.local/share/baloo/index lrwx------ 1 rrr rrr 64 5. úno 19.02 17 -> /home/rrr/.local/share/baloo/index lr-x------ 1 rrr rrr 64 5. úno 19.02 19 -> /usr/share/mime/mime.cache l-wx------ 1 rrr rrr 64 5. úno 19.02 2 -> 'pipe:[25294]' lrwx------ 1 rrr rrr 64 5. úno 19.02 3 -> 'socket:[27160]' lrwx------ 1 rrr rrr 64 5. úno 19.02 4 -> 'anon_inode:[eventfd]' lrwx------ 1 rrr rrr 64 5. úno 19.02 5 -> 'anon_inode:[eventfd]' lrwx------ 1 rrr rrr 64 5. úno 19.02 6 -> 'socket:[25908]' lr-x------ 1 rrr rrr 64 5. úno 19.02 7 -> /dev/urandom lrwx------ 1 rrr rrr 64 5. úno 19.02 8 -> 'socket:[25911]' lrwx------ 1 rrr rrr 64 5. úno 19.02 9 -> 'socket:[25912]' Baloo_file: ls -l /proc/3753/fd celkem 0 lr-x------ 1 rrr rrr 64 5. úno 19.05 0 -> 'pipe:[6806]' l-wx------ 1 rrr rrr 64 5. úno 19.05 1 -> /home/rrr/.xsession-errors lrwx------ 1 rrr rrr 64 5. úno 19.05 10 -> /home/rrr/.local/share/baloo/index lrwx------ 1 rrr rrr 64 5. úno 19.05 11 -> /home/rrr/.local/share/baloo/index lr-x------ 1 rrr rrr 64 5. úno 19.05 13 -> anon_inode:inotify lrwx------ 1 rrr rrr 64 5. úno 19.05 14 -> 'anon_inode:[eventfd]' l-wx------ 1 rrr rrr 64 5. úno 19.05 15 -> 'pipe:[25292]' lr-x------ 1 rrr rrr 64 5. úno 19.05 16 -> '/usr/share/mime/mime.cache (deleted)' lr-x------ 1 rrr rrr 64 5. úno 19.05 17 -> 'pipe:[25293]' lr-x------ 1 rrr rrr 64 5. úno 19.05 19 -> 'pipe:[25294]' l-wx------ 1 rrr rrr 64 5. úno 19.05 2 -> /home/rrr/.xsession-errors lr-x------ 1 rrr rrr 64 5. úno 19.05 23 -> 'pipe:[25296]' l-wx------ 1 rrr rrr 64 5. úno 19.05 24 -> 'pipe:[25296]' lrwx------ 1 rrr rrr 64 5. úno 19.05 3 -> 'anon_inode:[eventfd]' lrwx------ 1 rrr rrr 64 5. úno 19.05 4 -> 'socket:[7672]' lrwx------ 1 rrr rrr 64 5. úno 19.05 5 -> 'socket:[7673]' lr-x------ 1 rrr rrr 64 5. úno 19.05 6 -> anon_inode:inotify lr-x------ 1 rrr rrr 64 5. úno 19.05 7 -> /proc/3753/mounts lrwx------ 1 rrr rrr 64 5. úno 19.05 8 -> 'socket:[7680]' lrwx------ 1 rrr rrr 64 5. úno 19.05 9 -> /home/rrr/.local/share/baloo/index-lock No Baloo related information in .xsession_errors file. cat .config/baloofilerc [General] dbVersion=2 exclude filters=CMakeCache.txt,lzo,*.omf,CMakeTmp,confstat,moc_*.cpp,*.m4,*.elc,*.pc,*.o,*.aux,core-dumps,*.moc,cmake_install.cmake,libtool,*.la,config.status,.hg,.git,.xsession-errors*,*.po,.svn,ui_*.h,Makefile.am,autom4te,*~,*.rej,*.gmo,*.swap,.bzr,CTestTestfile.cmake,*.nvram,.pch,*.csproj,*.lo,CMakeTmpQmake,*.vm*,*.rcore,qrc_*.cpp,confdefs.h,po,*.loT,CMakeFiles,.obj,*.orig,*.class,lost+found,__pycache__,_darcs,conftest,CVS,.uic,*.part,*.pyc,*.tmp,.moc,.histfile.*,litmain.sh exclude filters version=2 first run=false How can I debug this please? Reproducible: Always
free total used free shared buff/cache available Mem: 8169780 685688 5291612 10956 2192480 6712044 Swap: 8396796 955516 7441280 Linux srv5 4.4.1-gentoo #1 SMP PREEMPT Wed Feb 3 10:46:09 CET 2016 x86_64 Intel(R) Core(TM)2 Quad CPU Q9550 @ 2.83GHz GenuineIntel GNU/Linux There is the index file of size 4446269440 Bytes present in .local/share/baloo after I killed the baloo_file_extractor process and stopped baloo using balooctl stop.
Same here, but I think it might be a 'balooctl status' display bug. It appears that the index is still there and searchable on my system: ~ $ balooctl status Baloo File Indexer is running Indexer state: Suspended Indexed 458732 / 506012 files Current size of index is 301.14 MiB ~/.local/share/baloo $ ls -al total 4502692 drwxr-xr-x 2 bjohnson users 4096 May 6 08:12 . drwx------ 46 bjohnson users 4096 May 5 14:57 .. -rw-r--r-- 1 bjohnson users 4610740224 May 6 08:16 index -rw-r--r-- 1 bjohnson users 8192 May 6 08:58 index-lock kinfocenter: Gentoo Linux KDE Plasma Version: 5.5.5 Qt Version: 5.5.1 Kernel Version: 4.4.6-gentoo OS Type: 64-bit Like David reports, once the index hits 4G 'balooctl status' reports 0MiB. However, search is still working on my system.
The reason for this bug is that the size is stored in a "quint32" variable, and the maximum value a quint32 variable can take is 2^32, which equals 4096 MB.
*** This bug has been marked as a duplicate of bug 364475 ***