Re: [Sshguard-users] what to do if sshguard goes down

[Mij <mi...@ss...>]

On Apr 8, 2010, at 8:46 , Adam Cohen wrote:

> occasionally i get a "real" crash and will report that when it happens next
> but the main thing that seems to take it down is when syslogd restarts
> 
> im running 1.4rc3 on Redhat with the "fifo" method, here's my startup command:
>   cat /var/log/sshguard.fifo | /usr/local/sbin/sshguard -a 3 -b  2:/var/db/sshguard/blacklist.db
> 
> whenever i see:
>   syslogd 1.4.1: restart.
> 
> on /var/log/messages, it is immediately followed by:
>   sshguard[pid]: Got exit signal, flushing blocked addresses and exiting...
> 
> makes sense why this happens, but how to restart?

Yes, intended/natural behavior.

I have a similar scenario, where I pass logs from one host to a Jail through a fifo/named pipe.

Inside the jail, a process (other than sshguard) gets logs from the fifo. I use indeed
daemontools to restart automatically the process when the fifo is closed/reset. "supervise"
avoid loops in case the other end of the fifo is not open as well.

http://cr.yp.to/daemontools.html