View a Journal's 'Specific Data' in understandable text

Glenda Miles

unread,

Dec 13, 2000, 3:43:44 AM12/13/00

to

Hi

I need to track what program is updating a specific field of a specific
file. This file and it's many fields are updated in numerous programs. Not
all the fields are always updated. The particular field I'm interested in
seems to be updated on a very sporadic basis. I have journaled the physical
file, but when I display the journal into an *OUTFILE the 'Specific Data' of
the fields is in either hex or EBCDIC I'm not sure but it's hieroglyphics to
me. Is there any way to get this data into a readable format so I can only
look at records where this particular field actually changed value.

Thanking you in advance.

Glenda

In for a penny

unread,

Dec 13, 2000, 4:05:31 AM12/13/00

to

Hi,

It is an image of the record format for the file. So just break it up
based on the file record format for the specific database file that the
journal entry is referring to.

Peter

============================

In article <3a37371e$0$2...@helios.is.co.za>,

Sent via Deja.com
http://www.deja.com/

Tony Goodwin

unread,

Dec 13, 2000, 9:43:47 AM12/13/00

to

Glenda,

I'll send you on some source(email direct as I dont want to post in
group) that I have created that
provides auditing of files at a field level without using jounalling,

Basically I use triggers on the insert/delete/update to record in an
audit file the previous and changed value (in readable form) job info,
date, time RRN etc. The audited fields are easily configured and I have
a prompt program to give subfile of changes for specific record and
field.

It is perhaps overkill for your purposes - but it will work.
You also could perhaps extract the logic to work on journal entries.

In article <917e4p$56d$1...@nnrp1.deja.com>,

--
-----------------
Tony Goodwin
mailto:tony.g...@dact.demon.co.uk

Charles Wilt

unread,

Dec 13, 2000, 10:48:54 AM12/13/00

to

Its pretty simple really, you need to build a temporary file whoose
structure is a duplicate of the journaled file with the addition of the
journal fields at the start of the file.

Now it would not be to difficult to design program that would do this on
the fly so you could use it for any file, you wouldn't even have to use
APIs to accomplish this although it would most likely be best if you did.
However, you may be able to find a shareware program that somebody else
has already created for this purpose, check http://www.as400network.com .

Lastly, there are commercial packages that include such functionality. I
use DBU from Prodata to handle this kind of stuff. One addtional feature
is the ability to search for a record.
Check http://www.prodatacomputer.com .

If you decide to roll your own, you'll want to start at the OS/400 Backup
and Recovery manual. It has a section about journaling you'll need to
know.

HTH,
Charles

In article <3a37371e$0$2...@helios.is.co.za>, gmi...@jba.co.za says...

Thomas

unread,

Dec 14, 2000, 12:33:18 AM12/14/00

to

Yes, all that's really necessary to view all of the entry in
human-readable form is the proper record format. The format can be
created through something such as interactive SQL by selecting all
fields (except JOESD) from probably QADSPJRN plus all fields from the
file being journalled. Use this to create the temporary file. Then do
the DSPJRN to an outfile and finally CPYF FMTOPT(*NOCHK) into your
temporary file.

The data is now in the temporary file and fully readable.

This can obviously be automated easily.

Tom Liotta

In article <MPG.14a163738...@netnews.worldnet.att.net>,

--
Tom Liotta, AS/400 Systems Programmer
The PowerTech Group, Inc.; http://www.400security.com
...and for you auto-things out there (NOT for people):
<a href="http://www.monkeys.com/cgi-bin/wpoison/wpoison.cgi">look!</a>