On Sun, 30 Nov 2014, Frédéric Brière wrote:

> Package: fail2ban
> Version: 0.9.1-1
> Severity: normal

> I've had fail2ban segfaulting yesterday (in libsqlite3; probably not
> worth a bug report, since I couldn't reproduce it).  Anyway, fail2ban is
> no longer running[*], so let's restart it:

Hi Frédéric,

yeah -- thanks for filing and confirming the issue on a sid box --
segfault issue is known to fail2ban upstream 
https://github.com/fail2ban/fail2ban/issues/844

and I have reported it against sqlite3 in debian (CCed with this as a
confirmation of replication on sid):
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=771341

There is some anecdotal evidence that recently merged upstream
https://github.com/fail2ban/fail2ban/pull/839
might somehow be a workaround for the segfault.

>   # service fail2ban start
>   [....] Starting authentication failure monitor: fail2ban[....] Socket file 
> /var/[FAILail2ban/fail2ban.sock is present ... failed!
>    failed!

> Yeah, I doubt it had the occasion to remove the socked file before
> segfaulting.  Nevertheless, the process is no longer running:

>   # kill -0 `cat /var/run/fail2ban/fail2ban.pid`
>   bash: kill: (24147) - No such process

> Shouldn't the init script automatically ignore/remove the socket file if
> the process in the PID file is no longer running?  Isn't that what PID
> files are for?  This is probably the first time I remember having to
> manually do something before restarting a server after a simple crash.
>  [*] Maybe this wouldn't be a problem if I was running systemd?  :)

this was intended as a "feature" -- fail2ban was so stable that abnormal
crash, leaving pid/socket behind, was treated as a warning sign that
something went really wrong and it requires admin attention to
troubleshoot/fix configuration etc.   There is a 'force-start' command
you could use to consciously ignore the fact that fail2ban has crashed.

-- 
Yaroslav O. Halchenko, Ph.D.
http://neuro.debian.net http://www.pymvpa.org http://www.fail2ban.org
Research Scientist,            Psychological and Brain Sciences Dept.
Dartmouth College, 419 Moore Hall, Hinman Box 6207, Hanover, NH 03755
Phone: +1 (603) 646-9834                       Fax: +1 (603) 646-1419
WWW:   http://www.linkedin.com/in/yarik        


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Reply via email to