zfs discuss - Aug 2008 - eWeek: corrupt file brought down FAA''s antiquated IT system

It is rare to see this sort of "CNN Moment" attributed to file
corruption.
http://www.eweek.com/c/a/IT-Infrastructure/Corrupt-File-Brought-Down-FAAs-Antiquated-IT-System/?kc=EWKNLNAV08282008STR4

 -- richard

On Thu, Aug 28, 2008 at 06:11:06AM -0700, Richard Elling
wrote:
> It is rare to see this sort of "CNN Moment" attributed to file
corruption.
>
http://www.eweek.com/c/a/IT-Infrastructure/Corrupt-File-Brought-Down-FAAs-Antiquated-IT-System/?kc=EWKNLNAV08282008STR4
`file corruption'' takes the blame all the time, in my experience, but
what does it mean?  It likely has nothing to do with the filesystem.
Probably an application wrote incorrect information into a file.

-- 
-Gary Mills-    -Unix Support-    -U of M Academic Computing and Networking-

On 28-Aug-08, at 10:11 AM, Richard Elling wrote:
> It is rare to see this sort of "CNN Moment" attributed to file  
> corruption.
> http://www.eweek.com/c/a/IT-Infrastructure/Corrupt-File-Brought- 
> Down-FAAs-Antiquated-IT-System/?kc=EWKNLNAV08282008STR4
>
"two 20-year-old redundant mainframe configurations ... that  
apparently are hanging on for dear life until reinforcements arrive  
in the form of a new, state-of-the-art system this winter."

And we all know that ''new, state-of-the-art systems'' are
silver
bullets and good value for money.

What goes unremarked here is how the original system has coped  
reliably for decades of (one guesses) geometrically growing load.

--Toby


>  -- richard
>
> _______________________________________________
> zfs-discuss mailing list
> zfs-discuss at opensolaris.org
> http://mail.opensolaris.org/mailman/listinfo/zfs-discuss

On 28-Aug-08, at 10:54 AM, Toby Thain wrote:
>
> On 28-Aug-08, at 10:11 AM, Richard Elling wrote:
>
>> It is rare to see this sort of "CNN Moment" attributed to
file
>> corruption.
>> http://www.eweek.com/c/a/IT-Infrastructure/Corrupt-File-Brought-
>> Down-FAAs-Antiquated-IT-System/?kc=EWKNLNAV08282008STR4
>>
>
> "two 20-year-old redundant mainframe configurations ... that
> apparently are hanging on for dear life until reinforcements arrive
> in the form of a new, state-of-the-art system this winter."
>
> And we all know that ''new, state-of-the-art systems'' are
silver
> bullets and good value for money.
>
> What goes unremarked here is how the original system has coped
> reliably for decades of (one guesses) geometrically growing load.
D''oh! It was remarked below the fold. I should have read page 2  
before writing.

The original architects seem to have done an excellent job, how many  
of us are designing systems expected to run for 2 decades? (Yes I  
know the cycles are shorter these days. If you bought a PDP-11 you  
were expected to keep it running forever with component level repairs.)

--Toby

zfs-discuss-bounces at opensolaris.org wrote on 08/28/2008 09:00:23 AM:
>
> On 28-Aug-08, at 10:54 AM, Toby Thain wrote:
>
> >
> > On 28-Aug-08, at 10:11 AM, Richard Elling wrote:
> >
> >> It is rare to see this sort of "CNN Moment" attributed
to file
> >> corruption.
> >> http://www.eweek.com/c/a/IT-Infrastructure/Corrupt-File-Brought-
> >> Down-FAAs-Antiquated-IT-System/?kc=EWKNLNAV08282008STR4
> >>
> >
> > "two 20-year-old redundant mainframe configurations ... that
> > apparently are hanging on for dear life until reinforcements arrive
> > in the form of a new, state-of-the-art system this winter."
> >
> > And we all know that ''new, state-of-the-art systems''
are silver
> > bullets and good value for money.
> >
> > What goes unremarked here is how the original system has coped
> > reliably for decades of (one guesses) geometrically growing load.
>
> D''oh! It was remarked below the fold. I should have read page 2
> before writing.
>
> The original architects seem to have done an excellent job, how many
> of us are designing systems expected to run for 2 decades? (Yes I
> know the cycles are shorter these days. If you bought a PDP-11 you
> were expected to keep it running forever with component level repairs.)
>
> --Toby
>
Then you also missed the all important crescendo where eweek uses the last
quarter of a poorly written article to shill completely unrelated but yet
inference to tie to the story software.

-Wade

On Thu, 28 Aug 2008, Toby Thain wrote:
>
> "two 20-year-old redundant mainframe configurations ... that
> apparently are hanging on for dear life until reinforcements arrive
> in the form of a new, state-of-the-art system this winter."
>
> And we all know that ''new, state-of-the-art systems'' are
silver
> bullets and good value for money.
The problem is that the replacement system is almost certain to be 
less reliable and cause problems for a while.  The old FORTRAN code 
either had to be ported or new code written from scratch.  If they 
used off the shelf software for the replacement then there is no way 
that the new system can be supported for 20 years.
> What goes unremarked here is how the original system has coped
> reliably for decades of (one guesses) geometrically growing load.
Fantastic engineering from a company which went defunct shortly after 
delivering the system.

Bob
=====================================Bob Friesenhahn
bfriesen at simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/
GraphicsMagick Maintainer,    http://www.GraphicsMagick.org/

On Aug 28, 2008, at 11:38 AM, Bob Friesenhahn wrote:
>  The old FORTRAN code
> either had to be ported or new code written from scratch.
Assuming it WAS written in FORTRAN there is no reason to believe it  
wouldn''t just compile with a modern Fortran compiler. I''ve
often run
codes originally written in the sixties without any significant  
changes (very old codes may have used the frequency statement,  
toggled front panel lights or sensed toggle switches ... but that''s  
pretty rare).



-- 
Keith H. Bierman   khbkhb at gmail.com      | AIM kbiermank
5430 Nassau Circle East                  |
Cherry Hills Village, CO 80113           | 303-997-2749
<speaking for myself*> Copyright 2008

On Thu, Aug 28, 2008 at 12:38 PM, Bob Friesenhahn <
bfriesen at simple.dallas.tx.us> wrote:
> On Thu, 28 Aug 2008, Toby Thain wrote:
>
> > What goes unremarked here is how the original system has coped
> > reliably for decades of (one guesses) geometrically growing load.
>
> Fantastic engineering from a company which went defunct shortly after
> delivering the system.


And let this be a lesson to all of you not to write code that is too good.
If you can''t sell an "update" (patch) every 6 months,
you''ll be out of
business as well :D

--Tim
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20080828/e4614133/attachment.html>

Tim,
> And let this be a lesson to all of you not to write code that is too 
> good.  If you can''t sell an "update" (patch) every 6
months, you''ll be
> out of business as well :D
Updates are generally functionality enhancements, which may well be 
chargeable; whereas patches are problem-fixes; which no-one will/should 
pay for... Call me pedantic ;-)

Regards... Sean.