Jean Francois Ortolo wrote:
> 
> On Sun, 23 Sep 2001, Martin Stricker wrote:
> >
> > root can do (nearly) anything. This is not Windows, in Unix there
> > is no method to *force* other processes not to touch a file. You
> > can ask them via fcntl() or flags in open(), but the other process
> > can just ignore this.

>   By "other processes", I believe you mean "processes specifically
> programmed to write in this library file". In such case, this would be
> some "hackerised processes", being made introduced in my system.

Not necessarily. Whichever procees damaged your libreadline did this
maybe in accident.

> Yes, but I've a good firewall, which denies all ports below 1024...

Weird firewall setup! Ports below 1024 are privileged, therefore any
intrusion needs a running service there which can be exploited. Ports
above 1024 are far more dangerous! Anything could sit behind such a
port.

>   In the meantime, there was no other process on my machine I could
> suspect even to use the libreadline.so.4.1 library file, apart from
> mysql itself. I made a 'rpm -q -R readline' recently, in order to show
> which were the other packages eventually able to use this library
> file. None of them was active at the mysql crash moment. ( don't
> remember exactly the list ).

A process using libreadline wouldn't write to it but only read it, so I
suspect a renegade process which was not aware there was anything on
that place of your harddisk. I think some process which had write rights
to libreadline did a random disk write out of accident. But without the
damaged file there is no way to know for sure.

> What happened?
>   Probably a sudden random libreadline.so.4.1 file change, apparently
> only due to hazard, and the harddisk behaviour. Weird.

Yes, it definitely is weird. But occasionally those things do happen,
and that's one of the reasons you keep backups. Like Michael Schwendt
pointed out those odd things are often the first sightings of growing
problems like hardware failure or a severe software bug. It could even
be a memory malfunction causing something to be changed into a write
instruction. Keep your eyes open, watch your logs and be happy if no
more weird things happen.

Best regards,
Martin Stricker
-- 
Homepage: http://www.martin-stricker.de/
Registered Linux user #210635: http://counter.li.org/



_______________________________________________
Seawolf-list mailing list
[EMAIL PROTECTED]
https://listman.redhat.com/mailman/listinfo/seawolf-list

Reply via email to