Page 2 of 2

Re: why we use keep in our jcl

PostPosted: Fri Nov 02, 2012 8:25 pm
by enrico-sorichetti
In my opinion IBM and the Navy are very well-managed organizations.

certainly not from the example You posted ...

zillions of better ways to protect things ...
hiding has never been the proper way of protecting things

Re: why we use keep in our jcl

PostPosted: Fri Nov 02, 2012 8:50 pm
by Akatsukami
c62ap90 wrote:Dick, since you cannot think of any regularly run Production Application jobs that would use NEW,KEEP,DELETE let me give you just 1 example from my experience.

When I worked for IBM I was contacted out to the Navy in San Diego. Some of the regularly run Production Application job datasets HAD to be NEW,KEEP,DELETE since they did not want anyone accidently keying the dataset name in, for example, 3.4 and pocking around these sensitive datasets.

You are aware, are you not, that if SMS is active on the system a disposition of KEEP will be overridden and CATLG substituted?

Re: why we use keep in our jcl

PostPosted: Fri Nov 02, 2012 9:06 pm
by c62ap90
Akatsukami wrote:
c62ap90 wrote:Dick, since you cannot think of any regularly run Production Application jobs that would use NEW,KEEP,DELETE let me give you just 1 example from my experience.

When I worked for IBM I was contacted out to the Navy in San Diego. Some of the regularly run Production Application job datasets HAD to be NEW,KEEP,DELETE since they did not want anyone accidently keying the dataset name in, for example, 3.4 and pocking around these sensitive datasets.

You are aware, are you not, that if SMS is active on the system a disposition of KEEP will be overridden and CATLG substituted?


I don't recall if the Navy had SMS back in 1992-4 when I worked there. ...or maybe SMS was not active for those KEEP datasets. I really do not know.

My original point is NEW,KEEP can be an option for some applications and since this is a forum for beginners and students it's worth learning - right? 99.99% of the time I use NEW,CATLG.

Re: why we use keep in our jcl

PostPosted: Fri Nov 02, 2012 9:13 pm
by dick scherrer
Hello,

Yes, they are mostly well-managed. And i doubt that a long-time IBMer sponsored this concept.

However, they (especially DoD) are also known for overkill - which is what this was. Being done in the name of "security" was intended to make it sacred and unquestionable. There are good ways to ensure security and others that only waste time, effort, money, etc. (Sorry, a personal rant<g>).

Having done work for all 4 branches of the DoD for more than 40 years, this was Never done for "extra protection" on any system i supported. Physical and system security protected datasets . . .

We actually had one fellow with all the "stuff" on his hat (1 star general) believe we could provide protection by putting this data in a "write-only" file. A colleague suggested this in jest and the general jumped right on it . . . A bit of an ugly time when it was explained to be a "funny" :)

Re: why we use keep in our jcl

PostPosted: Fri Nov 02, 2012 9:49 pm
by c62ap90
I agree with you 100% Dick!

Actually when I was with IBM in 1978 (geez I'm old) we used NEW,PASS a lot to save DASD space since it was very expensive back then. There was an extra charge for using CATLG. Of course space/DASD is cheap-ish now so NEW,CATLG is the norm.

Now I see a lot of CATLG datasets in shops that never seem to get touched/read after they are created and just get migrated. Lots of applications don't "cleanup" (ie. SHR,DELETE).

Too funny about the General.

Re: why we use keep in our jcl

PostPosted: Fri Nov 02, 2012 9:52 pm
by Akatsukami
c62ap90 wrote:My original point is NEW,KEEP can be an option for some applications and since this is a forum for beginners and students it's worth learning - right?

Maybe just barely, but probably not.

Talleyrand famously said of the restored Bourbon dynasts, "They have learned nothing and forgotten nothing". There are also "Bourbon analysts", who have learned nothing and forgotten since starting their careers 20, 30, or 40 years ago. As Mr. Scherrer says, they do not have many years of experience, they have one year of experience many times.

It is actually incorrect, IMNSHO, to say, "A disposition of KEEP means that the data set is to kept on the volume but not cataloged"; it is more accurate to say, "Historically, a disposition of KEEP meant that the data set is to kept on the volume but not cataloged; on modern systems with SMS is it the same as CATLG, unless the storage class allows the data set to be placed on a non-SMS-managed volume. Doing that without proper authorization, however, will get you fired and possibly blacklisted".

Re: why we use keep in our jcl

PostPosted: Fri Nov 02, 2012 10:38 pm
by c62ap90
So...
What you are saying is even Talleyrand would use NEW,CATLG ? Hee-hee.

Thanks.

Re: why we use keep in our jcl

PostPosted: Fri Nov 02, 2012 10:47 pm
by Akatsukami
c62ap90 wrote:So...
What you are saying is even Talleyrand would use NEW,CATLG ?

Talleyrand would probably bootleg the unused CPU cycles :P

Re: why we use keep in our jcl

PostPosted: Fri Nov 02, 2012 11:36 pm
by dick scherrer
Actually when I was with IBM in 1978 (geez I'm old) . . .


Not that old - as several of us can attest . . . ;)

d

Re: why we use keep in our jcl

PostPosted: Sat Nov 03, 2012 12:06 am
by Peter_Mann
Since we're walking down memory lane, and this thread has become somewhat off-topic, I remember working for a DOD contractor, security was tight, so tight that the operators kept a 3 ring binder with a list of dataset names and passwords to reply to at the console to allow certain job, users, access to files, these files were not only RACF protected but were contained in the PASSWORD dataset, talk about pariniod !
We also processes a lot of data thay enventually needed to go to each branh of the service, Navy, yup! only branch at the time still using 1600 bpi round tape, and we had to keel triple den tape drive around just for them, 1990's time frame, ah the good old days :)
have a good afternoon!