Re: Data conversion from 3380 to 3390 - H E L P

2006-03-23 Thread Ed Gould

On Mar 23, 2006, at 8:47 PM, Bruce Black wrote:

If you have DFDSS there is an option to reblock the datasets that  
are copied and its reasonably fast. I could be wrong in this but  
the FDR people didn't have the option.
Ed, luckily you are wrong.  Our dataset restore has the option to  
reblock PS and PO dataset.  Other dataset types like DA often have  
dependancies that prevent reblocking


Bruce,

Thanks. I Never saw this "option" going through the doc. I am happy  
to hear that it exists.


Ed



--
Bruce Black
Senior Software Developer
Innovation Data Processing

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Data conversion from 3380 to 3390 - H E L P

2006-03-23 Thread Bruce Black


Clearly the biggest issue is the geometry change from 47K to 56K; so you
can’t use any “physical” type utilities (E.g. Full Dump & Restore) to
perform the data move unless you configure the 3390 devices in 3380
emulation mode.  
Ya know, that is not completely true. As long as the output track will 
hold all the records on the input track, most access methods will 
support a track-by-track copy. VSAM usually will not, but does in some 
circumstances.


PO datasets are an issue because one of the quirky differences between 
3380 and 3390 is that the 3380 holds 1 more PDS directory block on a 
track than the 3390. So if the 3380 PDS has more than 1 track of 
directory (47 blocks I believe), then the track copy will fail


--
Bruce Black
Senior Software Developer
Innovation Data Processing

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Data conversion from 3380 to 3390 - H E L P

2006-03-23 Thread Bruce Black
If you have DFDSS there is an option to reblock the datasets that are 
copied and its reasonably fast. I could be wrong in this but the FDR 
people didn't have the option.
Ed, luckily you are wrong.  Our dataset restore has the option to 
reblock PS and PO dataset.  Other dataset types like DA often have 
dependancies that prevent reblocking


--
Bruce Black
Senior Software Developer
Innovation Data Processing

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Data conversion from 3380 to 3390 - H E L P

2006-03-22 Thread Ron and Jenny Hawkins
Patrick,

Most DBA would know this, so he can prepare a list and have you check it, or
vice versa. This is good old maker/checker, something that anyone that has
spent time in banking will be familiar with.

It's not like you are doing the job for the DBA - it is helping him. Would
you rather give him a list of files on 3380s, or hold hands at 4:30 in the
morning?

It's not like this is rocket science...

Ron

> -Original Message-
> From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On
> Behalf Of Mullen, Patrick
> Sent: Wednesday, 22 March 2006 11:37 PM
> To: IBM-MAIN@BAMA.UA.EDU
> Subject: Re: Data conversion from 3380 to 3390 - H E L P
> 
> If your Adabas DBA needs you to compile a list of the files that
> constitute the Adabas databases, then it's time to hire a new DBA. Or,
> at the very least, time to send them on a course.
> 
> 
> 

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Data conversion from 3380 to 3390 - H E L P

2006-03-22 Thread Mullen, Patrick
If your Adabas DBA needs you to compile a list of the files that
constitute the Adabas databases, then it's time to hire a new DBA. Or,
at the very least, time to send them on a course.



-Original Message-
From: IBM Mainframe Discussion List 

I agree. ADABAS is probably a better example than IDMS because IDMS
doesn't
have this problem.

Of course, if you know it is a problem for ADABAS you can use SMF to
compile
a list of all datasets that have been opened by ADABAS and give this to
the
DBA to use as a checklist.

Then you don't have to hold hands with the ADABAS DBA at 4:30 in the
morning...

Ron 

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Data conversion from 3380 to 3390 - H E L P

2006-03-22 Thread Fletcher, Kevin
Wally's purpose? To quote Wally, "That coffee cup does not get around by
itself."



That is just so Dilbert... What is Wally's purpose in life? Now why
didn't
we just do this for Y2K?

Thanks,
 
Fletch
 
 

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Data conversion from 3380 to 3390 - H E L P

2006-03-22 Thread Ron and Jenny Hawkins
Radoslaw,

I agree. ADABAS is probably a better example than IDMS because IDMS doesn't
have this problem.

Of course, if you know it is a problem for ADABAS you can use SMF to compile
a list of all datasets that have been opened by ADABAS and give this to the
DBA to use as a checklist.

Then you don't have to hold hands with the ADABAS DBA at 4:30 in the
morning...

Ron 

> The same apply to Adabas database datasets. Adabas uses their own access
> method (ADAM), which is geometry-specific.
> 
> In general I would say it is very good idea to *know* one's own
> datasets. Know means what is the purpose of given dataset, what
> application does use it, etc. Yes, it is possible even in case of
> hundreds thousands datasets.
> Otherwise - maybe half of them is just not needed anymore since 1989 ?
> 
> 
> --
> Radoslaw Skorupka
> Lodz, Poland
> 

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Data conversion from 3380 to 3390 - H E L P

2006-03-22 Thread R.S.

Ed Gould wrote:


On Mar 20, 2006, at 7:49 PM, Ron and Jenny Hawkins wrote:


SNIP---
You can use DCOLLECT to audit the attributes of your files. You can  
find and
DA files here and check if they need any special attention. You can  
also use
this DCOLLECT to identify your largest files. Just focusing on the  
largest

files that make up 80% of allocated space will get the most gain  for the
least effort.




Ron,

Good idea.. but not usually guaranteed. For some reason  (historical??) 
people who created BDAM files usually have dsorg=PS or  none specified. 
I have seen quite a few of those over the years  (including Panvalet 
libraries). I have not had the occasion to look  at IDMS databases 
lately but last time I did they were marked as PS.  So don't believe 
everything you see in the vtoc or SMF or decollect.


The same apply to Adabas database datasets. Adabas uses their own access 
method (ADAM), which is geometry-specific.


In general I would say it is very good idea to *know* one's own 
datasets. Know means what is the purpose of given dataset, what 
application does use it, etc. Yes, it is possible even in case of 
hundreds thousands datasets.

Otherwise - maybe half of them is just not needed anymore since 1989 ?


--
Radoslaw Skorupka
Lodz, Poland

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Data conversion from 3380 to 3390 - H E L P

2006-03-21 Thread Ed Gould

On Mar 21, 2006, at 6:45 AM, Ambat Ravi Nair wrote:


IDMS has been able to do VSAM as well ...


- ravi.



This was a while ago as the 3380 to 3390 conversion implies.

Ed

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Data conversion from 3380 to 3390 - H E L P

2006-03-21 Thread Ron and Jenny Hawkins
Ed,

That is just so Dilbert... What is Wally's purpose in life? Now why didn't
we just do this for Y2K?

Assuming you actually knew everyone that you would have to send an e-mail to
so your ass was well and truly covered, how would you know what to tell them
without actually doing any research?

Would you simply say "We are converting from 3380 to 3390 in 65 days. Please
convert all your applications by then." In the CYA realm that would be like
wiping your butt with confetti! 

If you don't have time for the luxury of research, why not hire a contractor
from WIPRO to do it. It would cost less than having a outage wouldn't it?

Ron

I'm sure you proudly put this migration technique in your resume!

> -Original Message-
> From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On
> Behalf Of Ed Gould
> Sent: Wednesday, 22 March 2006 3:57 AM
> To: IBM-MAIN@BAMA.UA.EDU
> Subject: Re: Data conversion from 3380 to 3390 - H E L P
> 
> On Mar 20, 2006, at 11:39 PM, Ron and Jenny Hawkins wrote:
> 
> > Ed,
> >
> > Agreed, and that is why SMF records are your second best friend :)
> 
> 
> ---SNIP---
> 
> Its always easy to sit back and spend the copious amounts of time
> researching for this stuff. I have never had the luxury of having the
> time. Sure some amount of research is needed but to spin through a
> years worth of SMF tapes was a luxury I did not have. I found if I
> knew a conversion was eminent (1-2 months) I would issue a memo and
> if they didn't correct the data sets I would move them anyway.
> 
> Sure IDMS wouldn't correctly come up but I had CYA'd myself. The
> people knew I didn't send out these memo's unless it was A. going to
> happen and  B. The problem would be their problem if the files
> weren't correctly identified and fixed before hand. C. The next time
> they would have to be in at 3AM with me to hand hold their special
> data sets.
> 
> That stopped a lot of griping and finger pointing.
> 
> Ed
> 

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Data conversion from 3380 to 3390 - H E L P

2006-03-21 Thread Ed Gould

On Mar 20, 2006, at 11:39 PM, Ron and Jenny Hawkins wrote:


Ed,

Agreed, and that is why SMF records are your second best friend :)



---SNIP---

Its always easy to sit back and spend the copious amounts of time  
researching for this stuff. I have never had the luxury of having the  
time. Sure some amount of research is needed but to spin through a  
years worth of SMF tapes was a luxury I did not have. I found if I  
knew a conversion was eminent (1-2 months) I would issue a memo and  
if they didn't correct the data sets I would move them anyway.


Sure IDMS wouldn't correctly come up but I had CYA'd myself. The  
people knew I didn't send out these memo's unless it was A. going to  
happen and  B. The problem would be their problem if the files  
weren't correctly identified and fixed before hand. C. The next time  
they would have to be in at 3AM with me to hand hold their special  
data sets.


That stopped a lot of griping and finger pointing.

Ed

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Data conversion from 3380 to 3390 - H E L P

2006-03-21 Thread Ambat Ravi Nair

IDMS has been able to do VSAM as well ...


- ravi.


Ed Gould wrote:

On Mar 20, 2006, at 7:49 PM, Ron and Jenny Hawkins wrote:
Ron,

Good idea.. but not usually guaranteed. For some reason (historical??) 
people who created BDAM files usually have dsorg=PS or none specified. I 
have seen quite a few of those over the years (including Panvalet 
libraries). I have not had the occasion to look at IDMS databases lately 
but last time I did they were marked as PS. So don't believe everything 
you see in the vtoc or SMF or decollect.


Ed


--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Data conversion from 3380 to 3390 - H E L P

2006-03-20 Thread Ron and Jenny Hawkins
Ed,

Agreed, and that is why SMF records are your second best friend :)

IDMS are marked as PS because they switch from BDAM to EXCP around 13 years
ago (Version 12, or maybe already in V10). I use to use DFSORT to split up
heavily accessed tables!

You can also use the 14/15 records to check the access method. If you find a
suspect access method then you can tie the 14/15 record to the type 30
subtype 4 record and get the program name.

Sure you won't catch everything, but you should find damn near almost
everything. 

Ron

> -Original Message-
> From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On
> Behalf Of Ed Gould
> Sent: Tuesday, 21 March 2006 11:49 AM
> To: IBM-MAIN@BAMA.UA.EDU
> Subject: Re: Data conversion from 3380 to 3390 - H E L P
> 
> On Mar 20, 2006, at 7:49 PM, Ron and Jenny Hawkins wrote:
> 
> 
> Ron,
> 
> Good idea.. but not usually guaranteed. For some reason
> (historical??) people who created BDAM files usually have dsorg=PS or
> none specified. I have seen quite a few of those over the years
> (including Panvalet libraries). I have not had the occasion to look
> at IDMS databases lately but last time I did they were marked as PS.
> So don't believe everything you see in the vtoc or SMF or decollect.
> 
> Ed
> 

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Data conversion from 3380 to 3390 - H E L P

2006-03-20 Thread Ed Gould

On Mar 20, 2006, at 7:49 PM, Ron and Jenny Hawkins wrote:

SNIP---
You can use DCOLLECT to audit the attributes of your files. You can  
find and
DA files here and check if they need any special attention. You can  
also use
this DCOLLECT to identify your largest files. Just focusing on the  
largest
files that make up 80% of allocated space will get the most gain  
for the

least effort.



Ron,

Good idea.. but not usually guaranteed. For some reason  
(historical??) people who created BDAM files usually have dsorg=PS or  
none specified. I have seen quite a few of those over the years  
(including Panvalet libraries). I have not had the occasion to look  
at IDMS databases lately but last time I did they were marked as PS.  
So don't believe everything you see in the vtoc or SMF or decollect.


Ed

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Data conversion from 3380 to 3390 - H E L P

2006-03-20 Thread Ron and Jenny Hawkins
Willie,

It's been a long, long time since I did this, but there are a few extra
weapons in the arsenal nowadays.

There is not much that is going to actually "break" when you go from 3380 to
3390. VSAM will look after itself and flat files will have wasted space. The
main things that will break are DA files.

First thing would be to think about your default device geometry in SMS.
Your allocation will be influenced by this. I would suggest leaving it as
3380 until your conversion is complete, and then change to 3390.

DCOLLECT is your friend. SMF records are your second best friend.

You can use DCOLLECT to audit the attributes of your files. You can find and
DA files here and check if they need any special attention. You can also use
this DCOLLECT to identify your largest files. Just focusing on the largest
files that make up 80% of allocated space will get the most gain for the
least effort.

Once you have found the largest files in DCOLLECT, use the Type 14, 15, and
6n records to find the job that allocates them. Replace the BLKSIZE parm
with the DSORG parm and they will be created with SDB. The small datasets
can be an exercise for a later time.

Ron

> 
> On Sat, 18 Mar 2006 04:56:12 -0800, willie bunter <[EMAIL PROTECTED]>
> wrote:
> 
> >Hi,
> >
> >  I am seeking help regarding the conversion from 3380 to 3390.  I
> performed this in 1992. I scanned the archives but there was a negligible
> amount of information.  Can anybody please help me?  Any advice will be
> greatfully appreciated.

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Data conversion from 3380 to 3390 - H E L P

2006-03-20 Thread Ed Gould

On Mar 20, 2006, at 6:38 PM, willie bunter wrote:

Thanks Jack for the suggestion.  Too bad the environment doesn't  
have HSM.  Your idea sounds like a great idea.  Out of curiosity  
what would happend if the files were to be recalled which volume  
would it be recalled to?  Would I need to keep the old volsers in  
order for the recall to work?



If you have DFDSS there is an option to reblock the datasets that are  
copied and its reasonably fast. I could be wrong in this but the FDR  
people didn't have the option.


Ed

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Data conversion from 3380 to 3390 - H E L P

2006-03-20 Thread willie bunter
Thanks Jack for the suggestion.  Too bad the environment doesn't have HSM.  
Your idea sounds like a great idea.  Out of curiosity what would happend if the 
files were to be recalled which volume would it be recalled to?  Would I need 
to keep the old volsers in order for the recall to work?

Jack Kelly <[EMAIL PROTECTED]> wrote:  if you have hsm up, an option is to 
migrate/recall and let hsm reblock (or 
just migrate). dss will also reblock src files so you could selectivly dss 
copy.

Jack Kelly
LA Systems @ US Courts
x 202-502-2390

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html




-
 Yahoo! Mail
 Use Photomail to share photos without annoying attachments.

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Data conversion from 3380 to 3390 - H E L P

2006-03-20 Thread Jack Kelly
if you have hsm up, an option is to migrate/recall and let hsm reblock (or 
just migrate). dss will also reblock src files so you could selectivly dss 
copy.

Jack Kelly
LA Systems @ US Courts
x 202-502-2390

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Data conversion from 3380 to 3390 - H E L P

2006-03-18 Thread Michael W. Moss
Hi Willie,

Clearly the biggest issue is the geometry change from 47K to 56K; so you
can’t use any “physical” type utilities (E.g. Full Dump & Restore) to
perform the data move unless you configure the 3390 devices in 3380
emulation mode.  The major issue will regard whether there are any
programs/JCL/CLIST,REXX et al that have hard-coded block sizes that are
device dependant.  This should be easily identifiable by scanning the
entire source portfolio, presuming you know where it all is.  If your shop
has stood still for a long-time and there really has been nothing but
running the workload that never fails, then configuring the 3390 devices
in 3380 emulation mode might be the way forward.

Most of use old-timers will remember going from 3330-3350-3380 and at the
3380 stage we probably decided enough was enough so we removed device
dependencies; but then along came DFSMS and 3390, the use of FBA RAID DASD
subsystems emulating CKD/ECKD and so we’ve never had another DASD geometry
change since then…

Best regards, UK Mikey.

On Sat, 18 Mar 2006 04:56:12 -0800, willie bunter <[EMAIL PROTECTED]>
wrote:

>Hi,
>
>  I am seeking help regarding the conversion from 3380 to 3390.  I
performed this in 1992. I scanned the archives but there was a negligible
amount of information.  Can anybody please help me?  Any advice will be
greatfully appreciated.
>
>  Thanks.

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Re: Data conversion from 3380 to 3390 - H E L P

2006-03-18 Thread Richards.Bob
Willie,

Either use SMS by allocating new 3390 pools to receive the NEW allocations and, 
as a result, do a lot of the migration by attrition.
Or use DFDSS/FDR to move the datasets logically to where you want them. 

If the 3380s are SMS managed, add 3390s to the pool and set the 3380s to 
DISNEW. 

A lot more info is needed for you to share with us, as an accurate answer 
depends on your current environment.


Bob Richards
VP, Enterprise Technologist
Enterprise Technology Infrastructure
SunTrust Banks, Inc.
(404) 575-2798 

Seeing beyond money (sm)

 -Original Message-
From:   IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED]  On Behalf Of 
willie bunter
Sent:   Saturday, March 18, 2006 7:56 AM
To: IBM-MAIN@BAMA.UA.EDU
Subject:Data conversion from 3380 to 3390 -  H E L P

Hi,
   
  I am seeking help regarding the conversion from 3380 to 3390.  I performed 
this in 1992. I scanned the archives but there was a negligible amount of 
information.  Can anybody please help me?  Any advice will be greatfully 
appreciated.
   
  Thanks.


-
Yahoo! Travel
 Find  great deals to the top 10 hottest destinations!

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html 
  
  
  
LEGAL DISCLAIMER 
The information transmitted is intended solely for the individual or entity to 
which it is addressed and may contain confidential and/or privileged material. 
Any review, retransmission, dissemination or other use of or taking action in 
reliance upon this information by persons or entities other than the intended 
recipient is prohibited. If you have received this email in error please 
contact the sender and delete the material from any computer. 
  
Seeing Beyond Money is a service mark of SunTrust Banks, Inc. 
[ST:XCL] 
 
 
 
 

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html


Data conversion from 3380 to 3390 - H E L P

2006-03-18 Thread willie bunter
Hi,
   
  I am seeking help regarding the conversion from 3380 to 3390.  I performed 
this in 1992. I scanned the archives but there was a negligible amount of 
information.  Can anybody please help me?  Any advice will be greatfully 
appreciated.
   
  Thanks.


-
Yahoo! Travel
 Find  great deals to the top 10 hottest destinations!

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html