Hi,

I'm running Firebird Superserver on Linux Debian 6.0 for a couple
of years.

Inspecting connected clients, I found a client process of my own (
a cron python script ) that didn't finished and remained hanged since
many days.

I killed it from linux shell ( killall <scriptname> ), and so when
I listed connected clients it effectively dissapeared.

Suddenly after, the firebird server remained deaf: it stopped
answering not only remote clients but local clients as well.
Running the command "isql-fb -user <username> -password <psw>
localhost:<databasename>" lasted forever.

Looking into /var/log/firebird2.5.log I found six of the following
messages:

<my_server_name> (Server) Mon Feb 10 15:01:26 2014
INET/inet_error: read errno = 104

I restarted Firebird server daemon and now everything seems to work
well ( I run a gfix validate to my database and it went OK )

As it is the first time in my experience that the firebird sever stops
responding, I'm concerned about what could be the cause of this
behaivour, and how to prevent it.

Thanks in advance for any diagnose.

Aldo Caruso