Thanks to:
treahy_at_allianceelec.com
alan_at_nabeth.cxo.dec.com
DAVID_PCOLAR_at_UNC.EDU
tpb_at_zk3.dec.com
flannon_at_odin.cair.du.edu
(Original question at bottom)
The easiest solution to this problem was to request our client to
resubmit the data dump in a Unix-readable format.
Thanks to several separate suggestions, I was finally able to view a
segment of the dump (past all the headers filled with garbage) toward
the end of the file (tail -50 13094_2303_h.2) and determine that the
right data was on the tape. However, the files still weren't readable
by our database load utilities. I am sure one of the following
suggestions would have worked. However all the suggestions were either
expensive or time-intensive.
Thanks again. The second time I asked fro help, I received numerous
responses. I shall learn to word my questions a little better!
Susan Rodriguez
susrod_at_hbsi.com
--------------
Now, it is remotely possible that IBM has utilities written for their
RS/6000 series of UNIX systems that can decode data from an AS/400, so
you might try doing some Web searching and asking IBM colleagues, if
you know any. But if you can't find something already written for a
UNIX system that does the work, then you're not likely to succeed if
the client doesn't have the technical expertise to help you out.
By the way, looking at some of the "plain text" in the original stuff
you had in your first message, it does look suspiciously like what I'd
expect for a raw dump of a relational database -- that is, descriptive
text in key places to facilitate debugging and processing. So you just
might be looking at the hardest possible problem.
Dr. Tom Blinn
--------------
This is a shot in the dark because my memory is vague at this point, but
a few years ago I had to dump an AS400 tape onto onto a sunbox. Anyway
the part I remember is that I used the dd command with the flag
conv=asci, since IBM stuff writes in ebcdic. Also you might have to
tweek the ibs (input block size) and the obs(output block size).
Hopefully you can find out from whoever wrote the tape what block size
was used when writing it.
Hope this helps.
Flannon
-------------------
Have you tried IBM technical support?
Look at
http://www.ibm.com, or call the local IBM sales office. They
can
probably assist reading the tape or pointing you to a provider that
will.
Good Luck.
Dave Pcolar
DAVID_PCOLAR_at_UNC.EDU
--------------------
did you try backup/list mua0:13094_2303h.2/save in the event that it is
a
saveset, but with an strange extention?
Have you tried to mount the tape foreign and use the DUMP command to
review
the first dozen or two records on the tape? It is possible that some
embedded ASCII text could identify the source of the tape, the technique
used to write the tape, etc... You would need to use the DUMP command
several times, just to get past the tape label information...
- - - - - - -
After I sent the message, it occured to me what this looked like. I
spent
my earlier years in IS shops, on older equipment, that spooled
journaling
files of database or application activities to tape. I suspect that
that
could be what you have here. Look at the label 13094_2303_h.2 could be
a
sloppy representation for the startup of this tape for
January 30, 1994 at 23:03pm
and the suffix.2 could be the second journaling tape of the set...
Barry Treahy, Jr. Email:
treahy_at_allianceelec.com
-------------------
ORIGINAL QUESTION:
----------------------------------------------------------
>> Hi,
>>
>> We are trying to load raw data files to an Alpha 1000A (DU3.2G) from
>> tapes dumped from an unknown system. The file names look a bit like
>> VMS: 13094_2303_h.2
>>
>> Question:
>>
>> Does anyone recognize this type of raw data and what type of
system/OS
>> it may have come from?
>>
>>
>> M-^?M-^?M-^?M-^?QSRDSSPC.1 ^YM-o^OM-^?^AM-<^C^A^V
>> ^V ^ZM-^_at_^AL/D OBJECT DESCRIPTOR
>> 3490#M-jM-.M-sHM-^DM-^?M-^H^AM-^_at_M-^X
>> ^A^DVQM-2M-#^ALM-\^D\^Z^Q\ ^DVQM-2M-#
>>
>> ^AM-^_at_aM-C^DVQM-2M-#^A^DVQM-2M-#^AM-C^YM-oQSRDSSPC.1
>> M-Q^ZM-^_at_M-^?^\^C^A#M-jM-.M-rM--GM-C^A^B^?^DVQM-2M-#^B^A#M-jM-.M-s
>> M-9$M-CM-^?^\3\3^AM-^ZMEMDTA
>> M-^\^A#M-jM-.M-raM-RM-C^A
>>
>> RECORD TYPE P1RTYPqP1DDTE
>> P1DDTE
>> M-^\^CM-^EM-^EM-^F^V-M-^Fk
>>
>>
>> DISCHARGE DATE P1DDTEqP1DISP
>> P1DISP
>> M-^\^CM-^OM-^O^B^V-M-^Fk
>>
>> DISCHARGE DISP P1DISPqP1EAUT
>> P1EAUT
>> M-^\^C^]^]^A^V-M-^Fk
>>
>>
>> Thanks,
>>
>> Susan Rodriguez
>> HBS International
>> Seattle, WA
Follow up e-mail:
>It's been 2 weeks, and I still have no idea how to get the data off my
>tape dump. Thanks to those few who ventured guesses as to the origin
of
>the data. After finally getting ahold of the people who sent it, I
know
>that the data came from AS/400. I can load the files to a ufs
>file-system on my Alpha running DU3.2G, but can only read junk from
them
>(see forwarded message).
>
>Is anyone aware of a way to read data files from an AS/400? Some of
the
>headers are readable (normally ascii characters), but there are a lot
of
>control characters and garbage interspersed. I assume I need some sort
>of translator/filter. Does such a thing exist?
>
>I'm about ready to send the tape back to the client and call it quits!
>
>susrod_at_HBSI.com
Received on Tue Sep 23 1997 - 22:27:15 NZST