RE: committment control and file changes -- MIDRANGE-L

Hi Joel,

Are the LFs being used in F specs in RPG?

If so, do they have the COMMIT keyword on the F spec?

If they do not, then that may be why changes will not roll back.

For commitment control to work in a native I/O environment all 3 of these things have to be in place:
1. The physical files involved have to be journalled
2. The job (or activation group) must be in commitment control, ie STRCMTCTL.
3. The files must be opened subject to commitment control, ie COMMIT keyword on the F spec. This applies to both physical files and logical files.

Regards,

Kevin Wright

LongReach - <http://www.lansa.com/longreach/> Share Files and Folders between IBM i Servers and iPhones/iPads <http://www.lansa.com/longreach/> - Download it for FREE today <http://www.lansa.com/longreach/> from the App Store <http://www.lansa.com/longreach/>

-----Original Message-----
From: midrange-l-bounces@xxxxxxxxxxxx
[mailto:midrange-l-bounces@xxxxxxxxxxxx]On Behalf Of Stone, Joel
Sent: Thursday, 31 May 2012 8:49 AM
To: 'Midrange Systems Technical Discussion'
Subject: RE: committment control and file changes

I agree, but I am arguing with other pgmrs here who swear
that chgs to LF's don't roll back, but chgs to the based-on
PF will roll back.

I would like someone on this BBS to explain how that is possible??

-----Original Message-----
From: midrange-l-bounces@xxxxxxxxxxxx
[mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of Alan Campin
Sent: Wednesday, May 30, 2012 5:25 PM
To: Midrange Systems Technical Discussion
Subject: Re: committment control and file changes

If a logical is just a SQL View and an SQL Index in one object and the
physical file it is based on is rolled back, why wouldn't the view and
index over the physical have the same values? It is just a projection
of the physical file.

On Wed, May 30, 2012 at 4:14 PM, Stone, Joel
<Joel.Stone@xxxxxxxxxx> wrote:

Yes, but...

If a PF is journaled, will a record update to a LF (based

on that PF) roll back under CM in the same manner that an
update to the PF would?

Thanks for response!

-----Original Message-----
From: midrange-l-bounces@xxxxxxxxxxxx

[mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of Vern Hamberg

Sent: Wednesday, May 30, 2012 5:00 PM
To: Midrange Systems Technical Discussion
Subject: Re: committment control and file changes

Joel

Have you looked at the help text of STRCMTCTL? It mentions some
requirements - among them, the file which will be processed under
commitment control must be journaled. There is NO

journaling of LFs - or

of OPNQRYF or anything other than PFs, as far as files go.

OPNQRYF is not a real file - it is a virtual access path,

one could say

- no data contained therein, only a PATH to the data, known

as an ODP or

open data path.

Likewise, LFs are only an access path.

You should also look at the STRJRNPF command - more

information there.

You should also look at the COBOL reference to see how to

specify using

commitment control on a file - in RPG it is on an F-spec -

I don't know

the equivalent.

It's all in the manuals at Infocenter - this is a pretty

big topic, much

going on around what might seem a simple question!!

HTH
Vern

On 5/30/2012 3:16 PM, Stone, Joel wrote:

Physical file changes under commitment control can be

rolled back for errors.

Do Logical files behave the same way ie the underlying PF

would roll back the same way that direct PF updates would?
Or are there some restrictions on LF's that would cause
programmers to use the LF to read the PF by primary key and
update the PF directly?

Do OPNQRYF's behave the same way as PF's also? Ie will

the based-on PF roll back on an error while under commitment control?

Does the file output method matter? Ie if a file is

opened as OUTPUT vs UPDATE would both of these behave the
same ? ie rollback on error under CC?

Some pgmrs here claim that PF updates roll back nicely,

whereas LF updates do NOT roll back. Is this possible?
(assuming all files are journalled properly and such)

Thanks!