I have some info for you.

Also, thank you for emailing me today about this. I did receive your letter.

I have some very important information on your hardware that I edited
using the online editor and for some reason did not save the many recent
additions.

I want you to know that I was online when the glitches began. I noticed everything
occur in short instances beginning with the alternate ftp server netftp that spread
onto intermittent t35 directory access issues. I noticed larger files would upload
and register yet not write on netftp.

I believe that there might have been a dma issue. or someone was trying to gain control
of the server and caused chmod issues.

This could be caused by hardware memory viruses or emf buildup along the raid array
through the chassis via dust fan blow by. Dust and fans dont mix. static can be created
by moving dust. Static electricity can power a home using just dust and steel or iron if you
move dust at a constant along metal. Anyway, I dont think dust hit your system.

On the other hand was the environment prone to humidity?

Also I noticed root directory access fairly upon using the t35 editor a few times. Access
privileges would fail 30 minutes after login despite periodic re-logins.

I'd login and upload edit things and upload others and within a short time I'd submit
changes and received a root directory access denial spanning until the system clock
turned over past the 12pm hour to the next day. sounds like something was set
temporarily and defaulted to a wait state as the day turned over on the system
clock.

This directory access obstructions occurred in time with the netftp write omissions..

In any event, data recovery should recover everything if the person knows what
their doing.

If they have to swap platters that may be the doable extreme. I think swapping the
drive boards should solve recovery access problems as partitions/geometry/bad
blocks might perform better.

All bad blocks have data on them that can be recovered if the tech can get to
them. Every recovery tech knows this.

Platter swaps take the same read time yet are risky because the techs lab might
not be sterile and he might not use a dust suit and gloves in a clean room.

There shouldn't be anything lost and I seriously believe you should ever need platters
changed. There wasnt a fire or explosion was there? No.

I really think chips were hit with signal interference via a hack or emf. Tiny zener
diodes that are meant to fail to protect circuits might also cause similar symptoms
that follows the circuits path through the board system diode to diode.

Constant build up of tiny amounts of emf can also cause an up and coming short circuit.
All this is speculation.

What is certain is that It would be best not to blindly accept the recovery technician's
word of any lost data because that could only happen upon a platter totally melting,
high energy degauss, hammer to the unit, complete low level format several times over or
similar. Even a low level format still leaves data in places that a recovery tech can revive.
Any total quick format is fully recoverable from despite the illusion of lost partitions
and master records wiped or block space corruption. Corrupted blocks still can be read
to some extent.

You can even recover data of of scratched or dent platters using the right technique.
If they say parts cant be recovered that means they don't know how to get them for
you. your still in the green.

Raid arrays fail in an intermittent stream all the time.

Also, can you have system logs for that hardware's local user access. It sounds to me
like a local user did something with the system that risked integrity.
Like say.........Shall we say.............Something like...Uh......................perhaps.........

..............LOGIN IN AS ROOT!!!!!!! [da.dah..dahhhhhhhhh!!!!!!!!]

All you have to do is log in and log out in some instances. If in windows then
the sky's the limit with what was going on. Everything happens to windows.

Never log in as root or wheel for anything unless the life of the system is screaming
for help. You logged in as root didnt you. Any how someone may have or someone
went to a girly site to check out some chicks and was hit with throughput
malware........

When its all said and done either its over exposed, slowly dying board issue or
its a programming/user failure issue.

Is your firewall script top notch. I got a huge one that might hit the spot.

When anyone logs in as root or administrator to any system linked to the public that system raises daemons that move and to their stuff instantly without user instruction.
these daemons are offered administrator privilege. when someone logs out errors
start developing system wide. The inclination is to at some point log in again as administrator and try to search for the problem. This raises open doors for everyone
who'd been scanning planting bots in the instance of eavesdropping hacker, or just
waiting for a chance to open public port access browse the internet in the instance
of a local user using the system like it was a their desktop. Open doors eventually get
larger.

Also did you write protect your hardware, and limit off platter access access only
certain nonessential parts. If so how? I think lighter rather than heavier security
is the most guilty culprit. Security means: Checking hardware for reliable
operability and migrating to new if needs be, setting pots and traps for snoops
to drop in, blocking risky processes from accessing system functions, keeping
a limit on who does what with what.....System admin 102. Dont be satisfied
with textbook advice because its might not be the only textbook.
Check the ibm.com website for administrator security tutorials.

Most important advice:

Find another 3rd raid array and disconnect if if the network and but in storage.
[grandfathering is the mother of all backup plans] redundant means spare has direct access to any system that might have a fault. Thus if faults have the ability to
communicate then the redundancy feels the strike also. Maybe adding an inline
power switch would have solved the problem. This should have been the first thing
you'd consider and if so why not? What might have been causing reluctance?
What if the system faulted in some other more critical way?

After I went through many of these I dropped Microsoft and dodged whatever
market mainstream I found myself within thereafter.

To be completely honest, the raid card has been giving us trouble for some time, but I kept thinking it was temporary and thought at worse we'd replace the card when it failed. I've never seen a raid card failure take out all the drives with it. And I have already taken your advice, we'll be doing massive backups of our other systems over the weekend and "putting them away" like you suggested

I actually ordered a new raid card, but it's a specific one we have been using (and would need that specific one to avoid hours of down time rebuilding the array). It was on back order and should be arriving in 2-3 weeks We actually ran a few diagnostic checks and everything came back OKAY. It was hard to tell if it was a hard drive issue or something else.. but with our raid setup and an extra batch of hard drives, it shouldn't have been a big deal.

Yes, IBM contractors make mistakes too, but not as many as hourly high-school-aged DC tech's This has just been a very unfortunate circumstance and we will obviously be taking steps to prevent this from happening in the future (including some of the tips you mentioned).

RAM = yummy. 

my website is bugged up. I have looked over all my codes and they work. and it shows fine on windows 2000, but not on xp service pack 2 (dont know about any other versions) or vista home pro. i have contacted 2 of my high school coder teachers, and they couldn't help. i think this might a problem with your software. some of my websites show good, then the others dont.

well, i can see it fine in mozilla firefox, but not in internet explorer < or=6. please tell me y. i looked online, no help. in a advanced book, no help.

is it possible for you to check my code for me. if so, tell me what is wrong.

ok. i have fixed the problem. it was where i located the style tag. but now thee cp i really slow all of the sudden. i have a fast computer, but it doesn't rule that out. i normally have no problems, but its getting to be over a minute in waiting time just for a simple save document command.