RAID failure (Was: RE: Cost of tape drives)

Dean,

RAID didn't work? Please explain. Was it a setup error, and you were not
really running RAID? Did you loose two drives in the array at the same time?
The CE couldn't pump either? Please provide us with the details. I have many
clients running RAID solutions on the 400 and I would like to know if there
is a problem.

Also, in BRMS, do you provide a list of libraries to backup, or a list or
libraries _not_ to backup? Just curious.

-Walden

> -----Original Message-----
> From: mcsnet!midrange.com!midrange-l-owner@Mcs.Net
> [mailto:mcsnet!midrange.com!midrange-l-owner@Mcs.Net]On Behalf Of
> DAsmussen
> Sent: Wednesday, January 21, 1998 7:28 PM
> To: MIDRANGE-L@midrange.com
> Subject: Re: Cost of tape drives
>
>
> Neil,
>
> In a message dated 98-01-21 12:16:13 EST, you write:
>
> <<snip>>
> >     The problem is, if the backup _isn't_ unattended, it often
> doesn't happen.
> >     Heck, over the holidays we lost a week's worth of eight developers'
work
> due
> >     to a head crash on our development machine.  Someone had taken
> our source
> >     library out of BRMS' save list because it "took too long to back up",
> DESPITE
> >     the fact that the backup was unattended.  Duhhhhh!
> >
> >             I presume this individual is now unemployed and living
> >  in a van down by the river ?   :-)
>
> You would be wrong, sir!  Typical "big company" politics -- nobody's
exactly
> sure who did it, nobody's exactly sure _why_ the crash occurred (three
> versions of the story so far -- and counting), we don't need mirroring
because
> we've got RAID, RAID didn't work because -- well it just didn't, the CE
> couldn't "pump" any of the drives because -- well he just couldn't.
>
> The sad thing is, there is a pervasive attitude of "well it's _JUST_ the
> Development box".  Unfortunately, source code "_JUST_" isn't transferred
to
> the Production box!  Nobody from operations checked the 7 _THOUSAND_ page
job
> log from the restore to see why it _was_ 7K pages (can you say "n security
> and/or data format changes occurred?  I _knew_ you could!), and we've been
> fixing things piecemeal for almost a month as we find them to be missing
or
> incorrect.
>
> Wanna hear something _REALLY_ scary?  BRMS failed on 12/13/97 and the
on-call
> tech walked operations through a manual backup.  As the latter wasn't a
"BRMS-
> tagged" backup, the tapes went to the scratch pile instead of off-site
storage
> and _just happened_ to have not been used.  Y2K has been our focus since
March
> '97, and our source library was excluded from BRMS in July...
>
> Of course, "it's _JUST_ the development system".  All 80Gb and four
> plants'-worth (totalling a little over $2B worth of annual production) of
> source code.  No need to audit _that_ quarterly for disaster procedures,
is
> there?
> <<snip>>
>
> Must _Just_ Be Me,
>
> Dean Asmussen
> Enterprise Systems Consulting, Inc.
> Fuquay-Varina, NC  USA
> E-Mail:  DAsmussen@aol.com
>
> "I thought that I was wrong once, but I was mistaken..." -- Anonymous
> +---
> | This is the Midrange System Mailing List!
> | To submit a new message, send your mail to "MIDRANGE-L@midrange.com".
> | To unsubscribe from this list send email to
MIDRANGE-L-UNSUB@midrange.com.
> | Questions should be directed to the list owner/operator:
david@midrange.com
> +---
>

+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to "MIDRANGE-L@midrange.com".
| To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---