[LUNI] My Smoothwall died, what should I do?

[Mark Martin]

I hope that Sten is right and that everything is okay on your system after 
fscking your filesystems.  But machines do not usually (and should not) 
spontaneously crash.  This is a complete guess but it sounds like it could be 
the beginning of a hard drive failure.  Hard drive failures are more common 
than other types of hardware failure and drives do not usually completely 
fail all at once but slowly degrade over time.  One possible symptom of a 
hard drive failure would be errors related to IO retries in 
/var/log/messages.  In any case, looking at /var/log/messages is the best 
place to look first to get hints about what might have gone wrong.

If you can't boot the machine from the hard drive, boot from a rescue disk 
like tomsrtbt or boot into rescue mode from a Red Hat or Mandrake 
distribution CD.  Then run fsck on the partition containing /var, manually 
mount the partition, and look at the messages file.

If a drive is failing, my advice is to install the drive as a secondary drive 
on another machine, mount whatever partitions you can at a temporary 
location, and copy whatever critical data you might have off of the drive.  
If a failing drive is the problem, you might not be able to copy data from 
all of the partitions, which means that you might lose some data.  Since the 
machine in question was a firewall machine, there might not be any critical 
data beyond possibly your firewall configuration files.

Once you have copied whatever data you can salvage from the failing drive, 
replace the drive and install the OS on the new drive.  When you have 
finished installing the OS, you can then copy whatever data you salvaged back 
onto the system.  Of course, you don't have to follow these directions in the 
order that I have given them.  If it is crucial to get your firewall back up 
as soon as possible, I would suggest replacing the drive, installing the OS, 
and configuring a new firewall first and then fooling around with recovering 
data later.

I hope that this helps.  Good luck.

Mark

On Friday 23 August 2002 12:10, russ arbuthnot wrote:
> Late last night, the Smoothwall machine which was running as the
> gateway/firewall/masquerading machine for my small business went kaput!