!!!data conversion problem !! help me
jooyoup
It is a big problem for us to make a conversion from IBM main fraim
data into unix oracle data. Actually we did not consider data
conversion progress as such a time consuming one.
Our C/S system consists of two unix ORACLE servers and our data
conversion procedure is explained below. Is there any point to improve
our work speed, please advise us....
1) IBM main fraim data -> UNIX machine through reel device
*** We could not help using reel device since there is no file
transfer program
from IBM into unix.
2) UNIX raw text -> dummy table in ORACLE
*** Using SQL*LOADER in this process, we have loaded data in parallel
option.
3) dummy table -> source table
*** Since there exist a munber of priority among tables, we had to
use dummy table to load data into really used tables. Most of time of
data conversion process is spent.
If you have a good idea, don't hesitate to reply us.
Thanks in advance.
guzzibill
On 08/17/96, jooyoup composed a News article about !!!data conversion
problem !! help me:
>It is a big problem for us to make a conversion from IBM main fraim
>data into unix oracle data. Actually we did not consider data
>conversion progress as such a time consuming one.
> Our C/S system consists of two unix ORACLE servers and our data
>conversion procedure is explained below. Is there any point to
improve
>our work speed, please advise us....
>
>1) IBM main fraim data -> UNIX machine through reel device
> *** We could not help using reel device since there is no file
>transfer program
> from IBM into unix.
In a similar situation, I have used FTP (File Transfer Protocol)
between an IBM mainframe (EBCDIC character set) and a DEC-Alpha C/S
system (ASCII character set). The FTP routines will handle the
conversion between character sets on the sender and the receiving
machines TRANSPARENTLY! The data was destined for loading into a
rather complex Oracle data base application as well.
>
>2) UNIX raw text -> dummy table in ORACLE
> *** Using SQL*LOADER in this process, we have loaded data in
parallel
>option.
Here you have to determine in which environment you are going to
"massage" the IBM data into the required Oracle table loader formats.
You basically have two choices...
1. Use the mainframe and your people experienced in that technology to
"convert" the existing data from the legacy system into Oracle
friendly table load parcels...then FTP them across and use SQL*LOADER
with triggers etc OFF to load the tables.
2. simply FTP the legacy data over in its existing formats and file
structures, then have staff familiar with the Oracle/Unix tools
reformat and splice the data into table load parcels and use
SQL*LOADER to populate the tables.
Which method you choose will depend upon how much effort is involved
in refomatting the legacy data into the newer Oracle data base tables
AND the skill sets of the Staff you have available in your project
team. We went with a mixture of both choices, with most of the effort
being concentrated on the mainframe side but certainly some "C"
programming was also required when loading tables with complex
inter-dependencies.
>
>3) dummy table -> source table
> *** Since there exist a munber of priority among tables, we had to
>use dummy table to load data into really used tables. Most of time of
>data conversion process is spent.
>
>If you have a good idea, don't hesitate to reply us.
>
>Thanks in advance.
>
--
Bill Scollard
Calgary, Canada
MotoGuzzi - California III || NeXT Turbo Colour
Software AG - Natural || OmniWeb & the NET
BackGammon / Poker / skiing || Marriage & Kids
=====> What more could there possbibly be!<======
John and Robin
A tool sold by Software AG would help. It can read most any source
(including ADABAS C) and output Oracle load files. The name of the tool is
Passport.
John
jooyoup <joo...@tyshl.co.kr> wrote in article
<321663...@tyshl.co.kr>...
> It is a big problem for us to make a conversion from IBM main fraim
> data into unix oracle data. Actually we did not consider data
> conversion progress as such a time consuming one.
> Our C/S system consists of two unix ORACLE servers and our data
> conversion procedure is explained below. Is there any point to improve
> our work speed, please advise us....
>
> 1) IBM main fraim data -> UNIX machine through reel device
> *** We could not help using reel device since there is no file
> transfer program
> from IBM into unix.
>
> 2) UNIX raw text -> dummy table in ORACLE
> *** Using SQL*LOADER in this process, we have loaded data in parallel
> option.
>
