Re: PF Compiled Files with Dictionaries vs. SQL-Created Tables

As much as I believe in good database design, our experience with
Referential Integrity on the AS/400 has not been good. Basically IBM's
implementation sucks.

We have had constant problems with database corruption which IBM seems
unable to fix. Just a constant problem. The internal tables get corrupted
and suddenly you cannot do alters or you cannot get to the tables. Om some
case, we can fix with RCLDBXREF but most of the time it does not see the
problem.

I just had to fix another one by saving the library, deleting the library
and then restoring the library. Luckily it was a development library. If it
had been a production, it would have been a disaster.

The other big problem we have had with RI is locks. Normally locks only
apply to the physical table you are updating but with RI suddenly you have
problems with any table that has RI to it. A record gets locked in a table
that has nothing to do with the table you are updating expect there is a RI
back to this other table.

Suddenly, you are getting crashes because RI cannot get to the other table.
This is just constant.

Anyway, just a warning.

On Tue, Feb 1, 2011 at 12:46 PM, Luis Rodriguez <luisro58@xxxxxxxxx> wrote:

Brian,

There is a lot to be said for DDL described tables against DDS-PF
definitions, mainly in the area of data validation and performance (no more
MCH1202 :- ) ).

Check out this IBM paper:


http://www-03.ibm.com/systems/resources/systems_i_software_db2_pdf_Performance_DDS_SQL.pdf

HTH,


Luis Rodriguez
IBM Certified Systems Expert — eServer i5 iSeries
--



On Tue, Feb 1, 2011 at 2:56 PM, Brian Piotrowski <
bpiotrowski@xxxxxxxxxxxxxxx> wrote:

Thanks for everyone's input. I guess the question now becomes do we just
abandon the SQL table creation in favour of PF-style definitions or not.

As

it stands right now, I would guess 90% of our tables are still in the
PF-style design. Converting the remaining 10% might be easier in the

long

run.

Anyhow, it's food for thought for us. We have a plethora of things to do
even before we consider this endeavour (such as converting all of our

8i/4i

dates/times to a proper timestamp), but it's good to know for the future.

/b;

-----Original Message-----
From: midrange-l-bounces@xxxxxxxxxxxx [mailto:
midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of Charles Wilt
Sent: Tuesday, February 01, 2011 2:04 PM
To: Midrange Systems Technical Discussion
Subject: Re: PF Compiled Files with Dictionaries vs. SQL-Created Tables

You can continue to use a "dictionary" file even with SQL created

tables...

Simply use the
create table MYNEWTBL as
(SELECT fld1, fld2 from DATADICT)
with no data
syntax.

optionally adding "copy options" such as
INCLUDING IDENTITY COLUMN ATTRIBUTES
INCLUDING COLUMN DEFAULTS
ect...

There's been articles done about this, google
SQL "Field reference files"

HTH,
Charles


On Tue, Feb 1, 2011 at 1:19 PM, Brian Piotrowski
<bpiotrowski@xxxxxxxxxxxxxxx> wrote:

Hi All,

For years (prior to my arrival), the programs written around here were

all based on physical file definitions. However, these definitions were
never really consistent - a field of one type may not be the same type

and

length in another file. We then began moving to creating tables with SQL
statements, keeping in mind the importance of creating fields that were

the

same type and length regardless of the table in which they were used.
However, this was cumbersome because we always had to keep a running

list

of field definitions along with their sizes and types.

We're now looking to move back to physical file definitions with a

twist

- all definitions will be contained in a single dictionary file that

lists

all of the fields, their types and their lengths. Even though we will

have

to do a lot of data table conversion, I am hoping this will allow us to
centralize the definitions and not have to worry that a field in one

table

may not be the same in another.

Has anyone had any experience similar to our current conundrum and if

so,

did you solve it using a similar method? If not, what worked best for

you?

Thanks!

/b;

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Brian Piotrowski
Assistant Mgr. - I.T.
Simcoe Parts Service, Inc.
Ph: 705-435-7814 x343
Fx: 705-435-5029
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
http://www.simcoeparts.com

Please consider the environment. Don't print this e-mail unless you

really need to.

The information contained in this communication is confidential and

intended only for the use of those to whom it is addressed. If you have
received this communication in error, please notify me by telephone

(collect

if necessary) and delete or destroy any copies of it. Thank you!

--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing

list

To post a message email: MIDRANGE-L@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.

--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing

list

To post a message email: MIDRANGE-L@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.

--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing

list

To post a message email: MIDRANGE-L@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.

--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.