SEO

I'm having trouble finding documentation for SCO UW. Can anyone point
me to a good resource so I can RTFM?

My goal is to come up with a disaster recovery script and to test it.
At this point, we have no backup/restore software and do not have a
budget for any.

How do I back up our system using the native SCO UX commands?

Regards,
Way out of my comfort zone

On Mon, 10 Dec 2007, mvsguy wrote:
> I'm having trouble finding documentation for SCO UW. Can anyone point
> me to a good resource so I can RTFM?

http://www.zenez.com/cgi-bin/ou8faq/faq


--
Boyd Gerber <gerberb@zenez.com>
ZENEZ 1042 East Fort Union #135, Midvale Utah 84047

mvsguy typed (on Mon, Dec 10, 2007 at 12:49:51PM -0800):
| I'm having trouble finding documentation for SCO UW. Can anyone point
| me to a good resource so I can RTFM?
|
| My goal is to come up with a disaster recovery script and to test it.
| At this point, we have no backup/restore software and do not have a
| budget for any.
|
| How do I back up our system using the native SCO UX commands?
|
| Regards,
| Way out of my comfort zone

Whats your time worth? Disaster Recovery is not a fun time.
There are 3rd party products out there that will make life
a lot easier for you. There are some folks that think they
are a waste of time and money, when you can roll-your-own.
Thats all true till you lose data, have a hard drive crash,
have data stolen, the list goes on.

My hats off to you for awareness of the potential problems.
When evaluating what your data is worth, a good reliable
time-tested backup and crash recovery product should be reconsidered.
They really are not that expensive considering their benefits.

www.microlite.com [ BackupEDGE ]
www.veritas.com
www.cactus.com [ LONE-TAR ]

- Jeff H
Disclosure: I'm the LONE-TAR guy, and still recommend any
3rd party solution over what comes with the OS.

> Whats your time worth?

My time? I have to show up at least 8 hours each day. I might as
well do some work...

I've cleverly disguised my handle to cover the fact that I'm an MVS
guy. Mainframes. I know how important DR is.
I've had a full on disaster caused by IBM applied hardware
maintenance. We only lost 18 hours of data. Later that month, we did
our first DR test. We did amazingly well.

Thank you both for trying, but buying software is simply not in the
cards.


Maybe I should ask a specific question. How do I go about using the
emergency tape/disks I've been making?

Regards,
MVSGuy

In article <20071210212934.GA21590@lonestar.cactus.com>,
Jeff Hyman <scolist@cactus.com> wrote:
>mvsguy typed (on Mon, Dec 10, 2007 at 12:49:51PM -0800):
>| I'm having trouble finding documentation for SCO UW. Can anyone point
>| me to a good resource so I can RTFM?
>|
>| My goal is to come up with a disaster recovery script and to test it.
>| At this point, we have no backup/restore software and do not have a
>| budget for any.
>|
>| How do I back up our system using the native SCO UX commands?
>|
>| Regards,
>| Way out of my comfort zone
>
> Whats your time worth? Disaster Recovery is not a fun time.
>There are 3rd party products out there that will make life
>a lot easier for you. There are some folks that think they
>are a waste of time and money, when you can roll-your-own.
>Thats all true till you lose data, have a hard drive crash,
>have data stolen, the list goes on.
>
> My hats off to you for awareness of the potential problems.
>When evaluating what your data is worth, a good reliable
>time-tested backup and crash recovery product should be reconsidered.
>They really are not that expensive considering their benefits.
>
>www.microlite.com [ BackupEDGE ]
>www.veritas.com
>www.cactus.com [ LONE-TAR ]
>
>- Jeff H
> Disclosure: I'm the LONE-TAR guy, and still recommend any
> 3rd party solution over what comes with the OS.
>

To echo Jeff's comments I had this scenario a few years ago.

I was called into a strange system, and prior to making
modifications I made a tape backup. Then, as I always do, I
checked to see if the backup was readable.

It was NOT.

Upon checking I found the tape drive was so dirty it would not
read anything.

After about 15 cotton swabs and alcohol the swabs started looking
clean.

I then insisted they get one of the commercial backup programs - at
that time it was split evenly between Backup-Edge and Lone-Tar

They thought it was expensive but I insisted.

A few months later they moved to a newer building and a newer
machine, and I transfered everything over.

Then the new machine started having failures. This was in the
early '386 days when memory card often had external cables to keep
the speed up instead of using the ISA bus.

Being able to load and start restoring the complete backup tape
after loading only 2 floppies - so the restore was starting in
under 5 minutes impressed the client.

He was even more impressed when I had to do it two more times that
evening - finally tracing the problems to the SCSI controller that
was mounted underneath the motherboard - some really strange
Italian design IMO [and saying that the older users know exactly
what 3 letter brand was on that machine].

Without the bootable floppies that contained all the drivers
for SCSI, Serial, etc. you would have had to install the OS,
configure the tape drives, configure the terminals or network
devices, etc.

So instead of going home at 3 or 4AM - we were both out of there
by 1030pm [we started at the close of the business day].

My charges >IF< he had not had the backup utilities would have
paid for the backup programs at least twice over.

That night the client became a true-believer and came to realize
that a good backup program does not cost, it pays.

If you decide to 'roll-your-own' and to make sure the backups are
truly reliable then you will probably spend more $$ of your time
than the program cost - unless you are paid less than a McDonalds'
employee.

Be aware the commercial programs perform bit-level compares on
the backup tapes/disks so that you KNOW the backup matches the
original exactly.

Be aware that just running checksums on backup media will NOT
guarantee that you have an accurate backup.

Though it is rare, if the data leaves the drive and gets corrupted
on the way to the tape controller the tape drive will write a
checksum which includes THE ERROR in checksum. So IOW all you have
done is verified vis the checksum that the data matches that
which was received by the controller and is not guaranteed to match
the disk.

So you need to evaluate the cost of your time for constructing
a backup program/script, the cost of potential data recovery costs
- or hiring an expert for handling things beyond your expertise -
versus the cost of lost data.

One of my first experiences with commercial backups goes back
to 1986 when Maynard used the TEAC [I believe that was the OEM]
data cassette and using Backup-Edge [though I think it had a
different name them]

There were two external tape drives and these were rotated on
schedule between 11 Xenix machines at the site.

Worked flawlessly except for the time on a Monday I restored
a Friday tape and NOTHING WORKED.

Unbeknownst to me, that office started shutting down early on
Friday, so they only backed up Monday thru Thursday, and the Friday
tape actually was about 1.5 years old and was for a machine
configuration far different that in use now.

I'd recommend seriously considering whether the cost of a
commercial backup will be that expensive in the event of a real
system crash, and the amount of downtime.

Remember that down time cost increases with the number of users
on a system. 100 users down for an hour could be thought of
as [approx] 100x $20/hr - or $2000 hour.

Bill
--
Bill Vermillion - bv @ wjv . com

In article <45ccd559-d45a-4a9b-8b4b-5bfa078be4af@f3g2000hsg.googlegroups.com>,
mvsguy <kkinney@fuse.net> wrote:
>> Whats your time worth?
>
>My time? I have to show up at least 8 hours each day. I might as
>well do some work...
>
>I've cleverly disguised my handle to cover the fact that I'm an MVS
>guy. Mainframes. I know how important DR is.
>I've had a full on disaster caused by IBM applied hardware
>maintenance. We only lost 18 hours of data. Later that month, we did
>our first DR test. We did amazingly well.
>
>Thank you both for trying, but buying software is simply not in the
>cards.
>
>
>Maybe I should ask a specific question. How do I go about using the
>emergency tape/disks I've been making?
>
>Regards,
>MVSGuy

Perhaps it >might< help if you described exactly how you made
the emerbency disks and tape?

Have you tried booting from the disk to read the tape?

That seems to me to be the first step.

Bill


--
Bill Vermillion - bv @ wjv . com

On Dec 10, 8:44 pm, mvsguy <kkin...@fuse.net> wrote:
> > Whats your time worth?
>

Hello mvsguy,

One last plug for the commercial stuff...

> My time? I have to show up at least 8 hours each day. I might as
> well do some work...

they help out on those days when you show up for work only to find
that you need 25 hours in the day to get the job done.

>
> I've cleverly disguised my handle to cover the fact that I'm an MVS
> guy. Mainframes. I know how important DR is.
> I've had a full on disaster caused by IBM applied hardware
> maintenance. We only lost 18 hours of data.

Do you mean 18 hours times a thousand users? Or 0.018 minutes per
user a thousand times? The former sounds catastrophic to me.
Likewise, if a drive fails and you can't boot the Unixware machine,
how
many users will be sidelined, and for how long, while you do the
recovery?

>Later that month, we did
> our first DR test. We did amazingly well.
>
> Thank you both for trying, but buying software is simply not in the
> cards.
>

Cactus and Microlite offer their prodcuts on a time-limited eval
basis.
Get the free download, create the emergency boot media and create a
master
backup. Lock them away in case disaster strikes. AFAIK, you can
restore from them even after the eval period has expired.

So far, you've spent only your time and no discretionary dollars.
Then, 6 months
from now, or whenever something bad happens, you can restore the
system to
the way it looks today. Then restore from whatever backups you create
on a
daily basis. Then, ask your boss if you can spend the ~$400 to buy
the software.

> Maybe I should ask a specific question. How do I go about using the
> emergency tape/disks I've been making?

Sorry, I don't know enough about Unixware to be specific. I'd start
by
testing the floppies, to make sure they can boot a computer other
than
the production system. Once you know the floppies work, schedule
a time when you can test boot the production system with them. Don't
assume that, because the floppies booted a test machine, they will
successfully boot your production system.

Once booted from floppies, test to see if you can mount the hard disk,
and read
your backup tape in its entirety (note that this does not verify the
data
on the tape). Duplicate the floppies. Don't forget to store an image
on the
production system (so they get backed up). Buy some disk drives to
use
as spares, preferably exactly what you have in the production box.
Booted
from the floppies, see if you can get your data onto one of those
spare hard
drives, and be able to successfully boot from that hard drive.
Carefully
document all of the steps, because that's what you'll be facing when
you
have a disk failure.

A better approach might to work from the commercial backup evals
again.
Test a disaster recovery to a spare drive. Choose the step by step
method
so that you can get an idea of all that's involved.

Best regards,
Dan Martin

>
> Regards,
> MVSGuy

In article <45ccd559-d45a-4a9b-8b4b-5bfa078be4af@f3g2000hsg.googlegroups.com>,
mvsguy <kkinney@fuse.net> wrote:

[second followup - wjv]

>> Whats your time worth?
>D
>My time? I have to show up at least 8 hours each day. I might as
>well do some work...

>I've cleverly disguised my handle to cover the fact that I'm an MVS
>guy. Mainframes. I know how important DR is.
>I've had a full on disaster caused by IBM applied hardware
>maintenance. We only lost 18 hours of data. Later that month, we did
>our first DR test. We did amazingly well.

>Thank you both for trying, but buying software is simply not in the
>cards.

IMO - as I mentioned in a prevous post - VERIFYING that the data
on the tape matches exactly is quite important. And actually
when you perform a restore, run the verify software to make
sure the data on the HD now matches the tape.

Years ago - one my Esix 5.3 [system] [I got tired of waiting for
the SCO 5.3] I ran a program call cktar - which acted remarkable
like the bit level verify on a program called C-tar that came out
a bit later.

If you want to put in a bit of effort search for the alt.sources
hierarchy [probably stored somewher] and look for cktar by
Wareen Tucker. It will be in the fall of 1990 archives, and
it will be about a 14K .c file tucked inside a shar archive.

I used to use it all the time when I was backing up to floppies,
as it would find an error every few disks.

Bill
--
Bill Vermillion - bv @ wjv . com

Bill Vermillion typed (on Tue, Dec 11, 2007 at 03:13:24PM +0000):
| In article <45ccd559-d45a-4a9b-8b4b-5bfa078be4af@f3g2000hsg.googlegroups.com>,
| mvsguy <kkinney@fuse.net> wrote:
|
| [second followup - wjv]
|
| >> Whats your time worth?
| >D
| >My time? I have to show up at least 8 hours each day. I might as
| >well do some work...
|
| >I've cleverly disguised my handle to cover the fact that I'm an MVS
| >guy. Mainframes. I know how important DR is.
| >I've had a full on disaster caused by IBM applied hardware
| >maintenance. We only lost 18 hours of data. Later that month, we did
| >our first DR test. We did amazingly well.
|
| >Thank you both for trying, but buying software is simply not in the
| >cards.
|
| IMO - as I mentioned in a prevous post - VERIFYING that the data
| on the tape matches exactly is quite important. And actually
| when you perform a restore, run the verify software to make
| sure the data on the HD now matches the tape.
|
| Years ago - one my Esix 5.3 [system] [I got tired of waiting for
| the SCO 5.3] I ran a program call cktar - which acted remarkable
| like the bit level verify on a program called C-tar that came out
| a bit later.
|
| If you want to put in a bit of effort search for the alt.sources
| hierarchy [probably stored somewher] and look for cktar by
| Wareen Tucker. It will be in the fall of 1990 archives, and
| it will be about a 14K .c file tucked inside a shar archive.
|
| I used to use it all the time when I was backing up to floppies,
| as it would find an error every few disks.
|
| Bill
| --
| Bill Vermillion - bv @ wjv . com

Bill,

You're really shaking my memory :-) Remembering back in the old
days when it was floppies for backup, even before Altos MTU4.
It was the "compression" that was the supertar magic, reducing 30 floppies
down to 3 or 4. It really was a big deal. The list goes on and its
interesting seeing how the industry changes, and is a moving target for
technology.

Hang in there,
Jeff H

In article <20071211182626.GA1091@lonestar.cactus.com>,
Jeff Hyman <scolist@cactus.com> wrote:
>Bill Vermillion typed (on Tue, Dec 11, 2007 at 03:13:24PM +0000):
>| In article <45ccd559-d45a-4a9b-8b4b-5bfa078be4af@f3g2000hsg.googlegroups.com>,
>| mvsguy <kkinney@fuse.net> wrote:
>|
>| [second followup - wjv]
>|
>| >> Whats your time worth?
>| >D
>| >My time? I have to show up at least 8 hours each day. I might as
>| >well do some work...
>|
>| >I've cleverly disguised my handle to cover the fact that I'm an MVS
>| >guy. Mainframes. I know how important DR is.
>| >I've had a full on disaster caused by IBM applied hardware
>| >maintenance. We only lost 18 hours of data. Later that month, we did
>| >our first DR test. We did amazingly well.
>|
>| >Thank you both for trying, but buying software is simply not in the
>| >cards.
>|
>| IMO - as I mentioned in a prevous post - VERIFYING that the data
>| on the tape matches exactly is quite important. And actually
>| when you perform a restore, run the verify software to make
>| sure the data on the HD now matches the tape.
>|
>| Years ago - one my Esix 5.3 [system] [I got tired of waiting for
>| the SCO 5.3] I ran a program call cktar - which acted remarkable
>| like the bit level verify on a program called C-tar that came out
>| a bit later.
>|
>| If you want to put in a bit of effort search for the alt.sources
>| hierarchy [probably stored somewher] and look for cktar by
>| Wareen Tucker. It will be in the fall of 1990 archives, and
>| it will be about a 14K .c file tucked inside a shar archive.
>|
>| I used to use it all the time when I was backing up to floppies,
>| as it would find an error every few disks.
>|
>| Bill
>| --
>| Bill Vermillion - bv @ wjv . com
>
>Bill,
>
> You're really shaking my memory :-) Remembering back in the old
>days when it was floppies for backup, even before Altos MTU4.
>It was the "compression" that was the supertar magic, reducing 30 floppies
>down to 3 or 4. It really was a big deal. The list goes on and its
>interesting seeing how the industry changes, and is a moving target for
>technology.
>
>Hang in there,
>Jeff H

You sent me email, to which I replied. It also wanted go to JPRs
site to get posted here, but I get refuse there.

You can post me reply if you'd like. I always prefer to answer
a post with a post and assume [incorrectly in this case] that the
email was just that - email and not for public reading.

Just color me 'an old fart' from days when DARPA was still part
of the 'net. It was surely nicer then with the ban on adversting.

Bill
--
Bill Vermillion - bv @ wjv . com