Monday bug triage

[Bryce Harrington]

Quite a few bugs this time, and a large chunk were mysql so I spent some
extra time there (more below):


LP: #1798089 - (New)            [clamav]         - ClamAV do not processes CL_TYPE_RAR cdb custom signature format 
  * Actually, "libclamunrar needs updated to 0.103.x to match clamav"
  * libclamunrar is in multiverse though, due to proprietary licensing.
    I retargeted the bug, and let reporters know this should rather be
    tackled in Debian.

LP: #1898928 - (Incomplete)     [ubuntu-advantage-tools] - package ubuntu-advantage-tools 19.6~ubuntu14.04.4 failed to install/upgrade: subprocess installed post-installation script returned error exit status 2
  * Users installed other distros on top of Ubuntu (Trisquel and Linux
    Mint) and lost their /etc/os-release
  * Marked bug WONTFIX and directed them to their respective distro
    maintainers.

LP: #1926265 - *(Confirmed)     [openldap]       - slapd enter in infinite loop on sched_yield syscall
  * Bug commenter confirmed reverting a patch fixes the issue
  * Not sure it is SRUable since the revert may regress a bug fix.
    So maybe needs a better fix developed?

LP: #1929270 - (New)            [mysql-8.0]      - package libmysqlclient-dev (not installed) failed to install/upgrade: libmariadb-dev:amd64 1:10.3.29+maria~bionic paketinde de bulunan '/usr/lib/x86_64-linux-gnu/libmysqlservices.a' dosyasının üzerine yazılmaya çalışılıyor
  * User filed a bunch of bugs against mysql-8
  * They attempted installing both mysql and mariadb, and each version
    of php since 5.6 via PPA.  Feels like they're setting up a server
    via cargo culted directions.
  * Recommend they not co-install mysql and mariadb, and closed all the
    bugs as invalid.

LP: #1929319 - (New)            [mysql-8.0]      - package mysql-server-8.0 8.0.25-0ubuntu0.20.04.1 failed to install/upgrade: installed mysql-server-8.0 package post-installation script subprocess returned error exit status 1
  * Looks like dupe of the above bug (timeout shutting down during
    upgrade)
  * However the log also shows a unique error message relating to locale
    settings that would be better to focus this bug report on.
  

In addition, I went through another half dozen recently reported mysql
upgrade bugs, suspecting there might be more bugs showing shutdown
failures coupled with long shutdown times.  Sure enough there are a
number of these.  In each case, previous shutdown/restart sequences took
only a matter of a few seconds, but the restart when the error happened
required over a minute.  Poking around in the /etc/init.d/mysql shutdown
code makes me suspect it has a timeout set to something less than a
minute (10 sec?) and then treats it as an error.  This might explain a
lot of the failures we see that seem to be complaining about existing
mysql services on other ports and so on.  This hypothesis should be
straightforward to test, and if it pans out the solution could be to
simply loosen the respective timeout to a larger number (maybe 5 min?)


Not all the mysql bugs were this same issue, there also continue to be
bug reports relating to co-installed mysql+mariadb.  Here is a way I
found helpful to examine what the user was doing leading up to the
error:

    $ grep "Requested-By: <username>" -B1 DpkgHistoryLog.txt  | grep Commandline

In examining a couple of these mariadb bugs this way, I get the sense
that the users may be installing mariadb semi-unintentionally by
cut-and-paste sysadminning and may be using online guides pertaining
more to Debian.

Bryce