I am looking for the internal logical format that IBM SAVOBJ save
files from IBM as/400's are written to tape.

My company receives this type of data on tape.  We can physically read
the data but need the internal format to parse it back into usable
data.  This internal format would be like the published specifications
of the Microsoft Tape Format.

The answer will need to give me the layout or logical map of how IBM
writes data to tape in the savobj format.

The correct answer will also get the maximum tip!!

---

Request for Question Clarification by denco-ga on 31 Mar 2006 18:32 PST

Howdy alex117-ga,

Do either of the following documents help?  Thanks!
http://publib.boulder.ibm.com/iseries/v5r1/ic2924/info/cl/savobj.htm
http://publib.boulder.ibm.com/iseries/v5r1/ic2924/info/cl/savobjsyn.htm

Looking Forward, denco-ga - Google Answers Researcher

---

Clarification of Question by alex117-ga on 03 Apr 2006 06:11 PDT

These links do nothing for me.  They are just the basic syntax of the
savobj command.  I am looking for the actual format of the data
written to tape.  Something like  byte 1-10 is this,  byte 11-13 is
this, and so on.

Howdy alex117-ga,

Do either of the following documents help?  Thanks!
http://publib.boulder.ibm.com/iseries/v5r1/ic2924/info/cl/savobj.htm
http://publib.boulder.ibm.com/iseries/v5r1/ic2924/info/cl/savobjsyn.htm

Can you upload a sample clip data and I will try to parse that.
I have done a lot of such work like COOEC.SUSPEND(a very old data
format which has no document) data parse.

http://www.rudas.at/Tips112000.htm

alex117

I am curious why you simply do not attempt to restore the SAVOBJ file
to an AS400 where it will be rendered in a more readable form? Even if
you do not possess an As400 (or the relevent tape drive) , there
should be plenty of  AS400 consulting firms willing to do this (for a
fee of course).

I am assuming the actual Algorithm that IBM uses to compact data( for
the SAVOBJ process) is something which they keep to themselves.

However another approach is if you say you can read the data would it
not be possible to write a program routine to restore the data to a
usable format?

Regards

Mongolia

http://www-900.ibm.com/cn/support/library/as400/download/HB548619.pdf

I thank everybody for their comments and links.  

First the links just have documentation for the actual command.

foxop-ga:  I will upload a small sample.  

mongolia-ga:  1.  We are not an as/400 shop and don't want to aquire
the expertise and start supporting IBM Systems.

2.  We get so many of these we want to do them in house.  (Right now
we do have some software the does does it, IMDRIVE and SHAFFSTALL but
they each of limitations.

3.  The algorithm to compact is not very complicated from what I understand.

4.  When you say another approach is..  That is exactly what I am
trying to do.  I just need to get the format.  Writting the program
will be the easy part.

alex117,
--I am not a Researcher--

forgive me if I misunderstand the problem but the capability to read
client data, from outside the OS/400 controlled environment, directly
from the client media  and write it to another (newer or different)
media in a different format might, if widely known, present a
potential security exposure.

A spectrum  of client scenarios:
An AS/400 installation wants to ship the offsite backups to a data
custodian with a cool, dry mineshaft ---> no reading required or
allowed.

An AS/400 installation wants a physically damaged tape to be restored
on a best-efforts basis by reading whatever is still retrievable  -->
can be done
(sometimes) by bit-for-bit duplication without needing to make sense
(parse) the *user* data.

An AS/400 installation outsources certain data management functions to
an agent whose lowest (best) bid won the business on the assumption
that "legacy" tape data could be read by RYO software utilities
running outside the OS/400 control program accessing used peripheral
drives.

---> IBM might license a utility that can accomplish the data
presentation while preserving client security and IBM proprietary
control data.


Again, forgive me if I misunderstand the problem description.