me? not exactly. it was aimed to be like that: if for some reason
socket is present while application is not running -- there could have
been some abrupt interruption of the process -- and the admin has to
check WTF and may be fix the cause, clean up leftovers of fail2ban's
operation (altered /etc/hosts.deny, iptables, etc) before attempting to
blindly start it again.
What I think I will do for the next debian release - I would
provide "force-start" action which would indeed use fail2ban-client -x
to overcome problem of the present socket file. but once again --
"start" action will remain as is
On Thu, 18 Oct 2007, Soren Hansen wrote:
> On Mon, Oct 15, 2007 at 09:48:07PM -0000, Nicolas Valcárcel wrote:
> > In this bug it would be better to patch the "/etc/init.d/fail2ban" or
> > make that on crash it remove the socket?
> Init-script definitely. If an application crashes, it's usually pretty
> difficult to tell it to clean up after itself. :)
> Will you create a patch for this?
--
.-.
=------------------------------ /v\ ----------------------------=
Keep in touch // \\ (yoh@|www.)onerussian.com
Yaroslav Halchenko /( )\ ICQ#: 60653192
Linux User ^^-^^ [175555]
--
fail2ban will not start if fail2ban socket is present
https://bugs.launchpad.net/bugs/123916
You received this bug notification because you are a member of Ubuntu
Bugs, which is a direct subscriber.
--
ubuntu-bugs mailing list
ubuntu-bugs@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
