[ale] Can't Boot! Need Super Block Wizard

Michael H. Warfield mhw at WittsEnd.com
Wed Jun 13 00:05:41 EDT 2012


On Tue, 2012-06-12 at 20:31 -0400, mdkliman at aol.com wrote:
> Scott,

> I'm grateful for your responses.

> I do not think the software broke.  Our problem is the result of
> repeated power and UPS failures.  I'm desperate for help in bringing
> the server back to life.

Concur...  You've suffered a hardware failure, based on your earlier
messages.  That error message for hda about "no media" is serious beyond
belief.  That indicates to me that you have a complete failure on that
drive, though the controller is still trying to talk to the interface.
I can't be sure if it's a motherboard problem, a power supply problem, a
drive controller board problem or and HDA (Head Disk Assembly).  All of
them are bad bad juju.

My next steps would be...

Try and power it down for a period of time to a complete cold stop and
then try and power it back up.  Listen to the drive very very carefully.
If you hear a faint buzz like a tone that buzzes and then stops and then
buzzes and then stops, the drive is probably trying to recalibrate to
cylinder 0 and you're looking at an HDA error.  If that's the case, the
drive is probably toast.

Take the faulted drive out and connect it to another system, either
mounting it in the system or in an external enclosure.  It's possible
you've had a partial power supply failure and the 12V spindle power is
dead.  It's also possible that the IDE interface on the motherboard has
failed (I've seen it).  Only way to really tell is to hook that drive to
another system and try to recover it.

If you can get it up, you can recover your data.  Best bet is that it's
a major failure on the old system and you better start looking at a new
system.  Most systems are seriously not worth repairing.  Even given
that a motherboard or power supply may be individually cheaper than a
new system, you won't know till you try it (and it doesn't work) and
then you still have a crappy old system that will fail again,
eventually.

If you can't get it up and your critical data is on that system, then
you have real problems.  Backups?  I'm not sure a recovery service will
be able to help with a "no media" error here.

I have used various voltages and temperatures to perform some miracle
recoveries but they are real "hail Mary pass" attempts that can take a
LOT of time (one effort at recovering a drive took over a month) with
zero guarantees.

You're saying "repeated power and UPS failures".  I hope you made some
really good backups.  Not to be judgmental here but, in the face of
that, I certainly would.  I would start to panic after the first such
failure with no good backups.

> Yes, this is an aged system that has had minimal maintenance for a few
> years.  It runs a dedicated-purpose application and we're careful to
> operate with in its limits.  I do not have the resources or the
> technical skills for upgrading or re-hosting this.

Oh, man...  You said Fedora Core 6?  That's like 5+ years old and 4
years past its end of life...  If you need a system that gets minimal
maintenance as a dedicated-purpose application server, you should NOT be
running Fedora.  I'm a Fedora fanatic and use it for everything but not
under those circumstances.  You need something with long term support
like CentOS / RHEL / SL.  You use Fedora if you want the latest and
greatest and are not afraid of investing the time and effort to keep it
up to date.

As far as upgrading and re-hosting - I don't think you're going to be
given much choice.  You've got a catastrophic failure on your hands.
First objective is to recover your data.  You may be able to "repair"
the system with either a new hard drive or a new power supply or a new
system entirely, but you're facing some work and probably more work to
do a repair in place than to migrate to a new, stable, system and
recover your data.

I don't think anyone can even begin to guess without a whole lot more
information...

How many drives?
Drive interface (IDE/ATA, SATA, SCSI, SAS)?
What size drive(s)?
Where's the critical database information?
Age and technical specs on the system?
Nature of application?
What DB (Postgres, MySQL, Oracle, other???)?

My instinctive remark is to say to boot this puppy up with a run-live
disk such as NST (Network Security Toolkit) and diagnose it from there
(just did one yesterday in an emergency) but that does take some
technical skills, which you say you don't have.

> Does anyone make house calls?

That could be a challenge, especially if this is time critical.  You say
you are in Norcross, which is close by but, I personally could not even
begin to look at it for at least another couple of weeks due to other
commitments.  Maybe someone else could be available but it sounds like
you need some pretty specialized skills and hardware knowledge.

> Marvin

Regards,
Mike

> -----Original Message-----
> From: Scott McBrien <smcbrien at gmail.com>
> To: Atlanta Linux Enthusiasts <ale at ale.org>
> Sent: Tue, Jun 12, 2012 8:14 pm
> Subject: Re: [ale] Can't Boot! Need Super Block Wizard
> 
> 
> As an aside, there is no such thing as Red Hat Fedora Core 6.  It's either Red Hat Enterprise Linux, which is a pay for product complete with support (to help with these types of issues), or Fedora Core 6, a community based distro that has about a 1 year life cycle.  If you're using Fedora Core 6 on production boxes (1) you're horribly out of date as they're now at Fedora 17 and (2) you shouldn't use community based distros for production systems.  Sure they're free, but you're on your own when stuff breaks.  If you *must* use a community based distro you want something like CentOS, which has a 7 or 10 year lifespan or Ubuntu LTS which has a 5 year lifespan.  If for no other reason than you can get updates (like to glibc which on your machine has several local user can take root exploits on it.)
> 
> 
> Perhaps we need a presentation on why one should choose an enterprise distro, as this is the second of these types of issues I've seen this week!  [the first was not an ALE reported issue :-) ]
> 
> 
> -Scott
> 
> On Jun 12, 2012, at 7:52 PM, mdkliman at aol.com wrote:
> 
> 
> 
> 
> I have a small business in Norcross, with 3 Linux servers running Red Hat Fedora Core 6.
> One of these is a database server running Oracle 11i.
>  
> We lost power and now the DB server won't boot.  It displays a message about not finding a "Super Block" and - I think - awaits entry of a utility name.  Our programmer is conversant with Linux, but told me I need help from an experienced Linux Admin.
>  
> "I've fallen and I can't get up."  I sure could use some help before I lose any more customers.
>  
> Thank you, in advance,
>  
> Marvin Kliman
> MTM Services
> cell:  (404) 433-0900  any time
> work:  (770) 441-3636 Ext 11  8am-6pm
>  
> 
> 
> _______________________________________________
> Ale mailing list
> Ale at ale.org
> http://mail.ale.org/mailman/listinfo/ale
> See JOBS, ANNOUNCE and SCHOOLS lists at
> http://mail.ale.org/mailman/listinfo
> 
> 
> 
> _______________________________________________
> le mailing list
> le at ale.org
> ttp://mail.ale.org/mailman/listinfo/ale
> ee JOBS, ANNOUNCE and SCHOOLS lists at
> ttp://mail.ale.org/mailman/listinfo
> 
> 
> _______________________________________________
> Ale mailing list
> Ale at ale.org
> http://mail.ale.org/mailman/listinfo/ale
> See JOBS, ANNOUNCE and SCHOOLS lists at
> http://mail.ale.org/mailman/listinfo

-- 
Michael H. Warfield (AI4NB) | (770) 985-6132 |  mhw at WittsEnd.com
   /\/\|=mhw=|\/\/          | (678) 463-0932 |  http://www.wittsend.com/mhw/
   NIC whois: MHW9          | An optimist believes we live in the best of all
 PGP Key: 0x674627FF        | possible worlds.  A pessimist is sure of it!
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 482 bytes
Desc: This is a digitally signed message part
Url : http://mail.ale.org/pipermail/ale/attachments/20120613/0e33149b/attachment-0001.bin 


More information about the Ale mailing list