https://bugs.kde.org/show_bug.cgi?id=517892

            Bug ID: 517892
           Summary: Kontact does not start: akonadi fails every few weeks
                    or months
    Classification: Applications
           Product: kontact
      Version First 6.6.1
       Reported In:
          Platform: Other
                OS: FreeBSD
            Status: REPORTED
          Severity: critical
          Priority: NOR
         Component: general
          Assignee: [email protected]
          Reporter: [email protected]
  Target Milestone: ---

Created attachment 190847
  --> https://bugs.kde.org/attachment.cgi?id=190847&action=edit
~/.local/share/akonadi/db_data/mysql.err

SUMMARY
Every few weeks or months, Kontact fails to start because the Akonadi DB does
not start.
I know we should be kind and friendly when writing bug reports, but in this
special case, I refuse to do so, because this bug exists for years and hasn't
been fixed and I have reported it numerous times!  Access to e-Mail & PIM in
general is CRITICAL!!!  Which <censored> is responsible for this piece of shit?
 Get him/her OUT!!!  Akonadi MUST be replaced with a ROBUST & RELIABLE DB,
ASAP, yesterday!  Up to now, the fix has always been to manually create a
missing file or directory '#xyz...'in ~/.local/ share/akonadi/db_data, I don't
remember the exact name (the redo log?).  But today the errlog is slightly
different.  Please note that I'm using ZFS, a copy-on-write filesystem with
strong checksums, so a failure on that level is mathematically IMPOSSIBLE, even
if the underlying hardware were corrupt (but it is not, and that would be
strange 'cause it's new).

Fortunately I have regular snapshots, so I'll be able to restore the last
working DB from 2 days ago, but it would be much better if the error messages
include an easy to understand receipe of how to fix it, because not anyone has
backups or snapshots.  Note: I'm a very experienced hardcore computer/IT nerd
for decades myself, but the error log entry "No valid checkpoint found
(corrupted redo log). You can try --innodb-force-recovery=6 as a last resort."
is not helpful, because it does not tell me WHERE to specify this parameter.

STEPS TO REPRODUCE
1.  Have FreeBSD 14.4-RELEASE installed on a Thinkpad T450, no special tweaking
in loader.conf(5), rc.conf(5) and sysctl.conf(5), just what the package
messages tell you during their installation.
2. Install KDE from the binary packages with pkg(8), i.e. do not build it
yourself from ports(7).
3. Use KDE & Kontact daily, and enjoy a smooth ride in general (except: see my
other bug reports).  Eventually, Kontact (Akonadi) will fail as described.

OBSERVED RESULT
Akonadi fails to start every few weeks or months, blocking desktop assess to
PIM with Kontact.  Then the user must access the PIM by web interface with a
web browser, with much less comfort. 
That the error message of the GUI is not qualified is another bug that I
already reported and will do again in a few minutes.

EXPECTED RESULT
PIM suite should work reliable 24x7 with industry grade failure rate of much
less than 0.001%.
Before even thinking of implementing new features, first fire 75% of the
current developers and let only those who know how to design & write software
do the work.  Development speed will improve drastically, because those 75%
<censored> do more harm than good.

SOFTWARE/OS VERSIONS
Kontact: 6.6.1 (25.12.1)
KDE Frameworks: 6.22.0
Qt: 6.10.1 wird verwendet und gebaut wurde mit 6.10.1
FreeBSD 14.4-RELEASE (Xcb)
Build ABI: x86_64-little_endian-lp64
Kernel: freebsd 14.4-RELEASE

ADDITIONAL INFORMATION

-- 
You are receiving this mail because:
You are watching all bug changes.

Reply via email to