> 1. How long does it take to make a tape backup of a 32760 cylinder
volume?

We have 1.3 TB allocated on Production, Our weekly full volume backups
using FDR starts Saturday evening 18:00, and runs to 4 drives (ESCON)
that creates the master and the duplicate at the same time 2x2, and it
runs for 8 / 9 hours now that we have an ATL that does not have to wait
for ops to load the tapes. Also we use Green tapes, which means they do
not have to fill a lorry load to send it off to DR site. I am sure if we
weren't shutting down, we would have received the money to get FICON,
which would have improved the time. Most of our restores comes from an
incorrectly configured FDR migration tool, that migrates the datasets,
just before someone needs it again, sad, but we have changed it a few
times but just haven't reached that optimum date that will minimize the
restores, Luckily BOB, the robot is not complaining, and his hand is too
new to have him dropping the tapes, so there is really no problem. On
all the other points, I fully agree with Jim.

Regards

Herbie

-----Original Message-----
From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On
Behalf Of Jim Phoenix
Sent: 09 Februarie 2008 02:20
To: [email protected]
Subject: Re: VTOC size

1.  Slightly less time than it takes to backup 32760/3339 (approx. 9.81)

mod 3 volumes that are being stacked to the same output tape.  Figure 
some housekeeping occurs when the backup software finishes one volume & 
then starts another.  The difference should be minuscule when total 
number of unused tracks is held constant.
2.  Modern tapes can hold multiple volumes, large or small.  They also 
record in a serpentine fashion, back & forth, so high speed positioning 
to a given location is quicker.
3.  From a dataset level backup, which is what should be used for this 
purpose, cylinder location is irrelevant.  Competent restore software 
should be able to restore a single dataset even from a full volume 
backup.  If the actual software in use cannot do this, and you don't 
have a dataset level backup available, then yes, restoring an entire 
large volume for a single dataset will take longer.  Note that the 
sooner the accidental modification is detected, the more likely it is 
that the changed data has not even been backed up yet, the likelier the 
latest dataset backup is valid.  Simple, quick restore.  For deleted 
datasets,  the recovery could be done months later, depending on your 
deleted dataset backup retention policy.

Skip Robinson wrote:
> The idea of massive volumes is intriguing, but my concern would be
> backup/restore. On some our non-z platforms we have huge volumes many
times
> the size of 3390-3. When one of them went south a while back, it took
hours
> and hours to get it restored. With RAID arrays, of course, that's not
> supposed to happen; I think it was a problem with an SVC.
>
> On the mainframe we take frequent full volume backups. The likelihood
of
> true failover (DR) is minuscule, and we use XRC mirroring for that
> contingency anyway. On the other hand the probability of having to
retrieve
> an accidentally deleted or modified file is huge--it happens all the
time
> even for sysprog finger checks.
>
> 1. How long does it take to make a tape backup of a 32760 cylinder
volume?
> 2. Where does that backup live?
> 3. How long does it take to restore a critical 10 track file from two
weeks
> ago that happens to live on the corner of cylinder 31753?
>
> .
>
>   

-- 
| Jim Phoenix                      | Voice:   (310) 338-0400 x316   |
| Senior Software Developer        | Fax:     (310) 338-0801        |
| Phoenix Software International   | Alt fax: (310) 337-2685        |
| 5200 W. Century Blvd., Suite 800 | [EMAIL PROTECTED] |
| Los Angeles, CA 90045            | http://www.phoenixsoftware.com |

Opinions expressed by this individual are not necessarily those of the
Company.

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
Elavon Financial Services Limited
Registered in Ireland: Number 418442
Registered Office: Block E, 1st Floor, Cherrywood Business Park, Loughlinstown, 
Co. Dublin, Ireland
Directors: Robert Abele (USA), John Collins,  Terrance Dolan (USA),  Pamela 
Joseph (USA), Declan Lynch, John McNally, Malcolm Towlson
Elavon Financial Services Limited, trading as Elavon, is regulated by the 
Financial Regulator

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Reply via email to