View previous topic :: View next topic
|
Author |
Message |
Anuj Dhawan
Superior Member
Joined: 22 Apr 2006 Posts: 6250 Location: Mumbai, India
|
|
|
|
Hi,
Problem: How to find which DSN belongs to which 'team' ?
Scenario: Need to delete/minimize the use of datasets which starts with some HLQ.
My approach:
1. List all the datasets starting with some HLQ.
2. I can use LISTCAT LVL(HLQ) ALL to do this, but for me the output is quite nasty, when I'm just intrested to look for only data set names. So I used IKJEFT1A, this solves my problem a bit.
3. Now that some HLQ is used across teams, how to find which DSN belongs to which team here at my shop ?
My senior suggests that create an excel of DSNs starting with that HLQ & send across the team to check which DSN belongs to which Team, then wait for their mail. Ah..it's a donkey-work.
I'm thinking to use SMF records, something on the line of this topic..
www.ibmmainframes.com/viewtopic.php?t=29449&highlight=
So I need two suggestions:
1. Is there any alternate way, as I'm thinking to use SMF records, to do this or not?
2. If we can use SMF records, then I'm stuck up here in how to create SMF Data. looked here
publib.boulder.ibm.com/cgi-bin/bookmgr/BOOKS/IEA2G281/2.3.1?DT=20080122045841
but couldn't get a good solution. Please suggest. |
|
Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006 Posts: 19244 Location: Inside the Matrix
|
|
|
|
Hello,
Quote: |
how to find which DSN belongs to which team here at my shop |
From smf you might find who (which teams) use each dataset name, but that may not identify which dsn belongs to which team.
One thing you might identify is a list of datasets that have not been referenced for some amount of time - a year, 2 years, 6 months. . . If you circulate a list and want responses, the way that i believe works best is to tell people that "the following datasets will be deleted on mm/dd/yyyy unless you send a reply that the dataset needs to be retained. Before doing anything, get a couple of backups of the potential scratch datasets. If "they" are not required to respond, many will not. . .
Quote: |
I'm stuck up here in how to create SMF Data |
I suspect that you will want to read the existing smf data rather than create new smf data. |
|
Back to top |
|
|
enrico-sorichetti
Superior Member
Joined: 14 Mar 2007 Posts: 10872 Location: italy
|
|
|
|
IDCAMS DCOLLECT |
|
Back to top |
|
|
expat
Global Moderator
Joined: 14 Mar 2007 Posts: 8797 Location: Welsh Wales
|
|
|
|
Yeah, DCOLLECT is not too bad to use and does contain a whole lot more information than LISTCAT.
What about naming standards, as at most shops I've been involved in the naming setup/revamp the HLQ was DEV and the 2LQ was project id, so that makes it easier to identify which group is eating up all the space.
How about migrating datasets to ML2 and then seeing which ones get recalled. Not the most efficient I know, but it can give a good idea of what's going on |
|
Back to top |
|
|
Anuj Dhawan
Superior Member
Joined: 22 Apr 2006 Posts: 6250 Location: Mumbai, India
|
|
|
|
Hi,
expat wrote: |
What about naming standards, as at most shops I've been involved in the naming setup/revamp the HLQ was DEV and the 2LQ was project id, so that makes it easier to identify which group is eating up all the space. |
Yeah, nearly same way it's here, but for that matter I've to be involved with all the teams & that makes this process 'manual' instead I'm thinking to do it using some JOB, any suggestion for that.
Quote: |
How about migrating datasets to ML2 and then seeing which ones get recalled. Not the most efficient I know, but it can give a good idea of what's going on |
I think, as an Application Engineer I'm not autorized to do it, I need to contact my system guys? |
|
Back to top |
|
|
skkp2006
New User
Joined: 14 Jul 2006 Posts: 93 Location: Chennai,India
|
|
|
|
Quote: |
I think, as an Application Engineer I'm not autorized to do it, I need
to contact my system guys? |
Can you just try the following...
HMIGRATE Command
You use this command to manually migrate a data set. You may want to do this a 'quick fix' to resolve space problems. At its simplest, the 3.4 line command is
HMIG DSNAME(/)
or
HMIG DSNAME(/) ML2 |
|
Back to top |
|
|
Anuj Dhawan
Superior Member
Joined: 22 Apr 2006 Posts: 6250 Location: Mumbai, India
|
|
|
|
Hi,
This
works for me without (/). However, would like to know can we use it in Batch? |
|
Back to top |
|
|
skkp2006
New User
Joined: 14 Jul 2006 Posts: 93 Location: Chennai,India
|
|
|
|
Anuj,
Yes, you can run the HSM Dataset Level Command HMIG In batch.
Try this...
Code: |
//STEP01 EXEC PGM=IKJEFT01
//SYSPRINT DD SYSOUT=*
//SYSTSPRT DD SYSOUT=*
//SYSTSIN DD *
HMIG 'SYAM.TEST.FILE1' ML2
/*
// |
I believe that you have'nt lost your memory... ;-)
ibmmainframes.com/about23202.html
Syam |
|
Back to top |
|
|
expat
Global Moderator
Joined: 14 Mar 2007 Posts: 8797 Location: Welsh Wales
|
|
|
|
Unfortunately only if he has ALTER access on the datasets being migrated. |
|
Back to top |
|
|
expat
Global Moderator
Joined: 14 Mar 2007 Posts: 8797 Location: Welsh Wales
|
|
|
|
You are obviously suffering from severe mainframe addiction |
|
Back to top |
|
|
expat
Global Moderator
Joined: 14 Mar 2007 Posts: 8797 Location: Welsh Wales
|
|
|
|
Quote: |
... ..I'm drinking, eating..& chewing Mainframes...hfff...mercy...God. |
Look on the bright side, it could be UNIX |
|
Back to top |
|
|
Bill Dennis
Active Member
Joined: 17 Aug 2007 Posts: 562 Location: Iowa, USA
|
|
|
|
anuj,
If all you want are names, use
Code: |
LISTCAT LVL(HLQ) NAME |
instead of the ALL parameter. The listing is way fewer lines! |
|
Back to top |
|
|
sameer
New User
Joined: 10 Mar 2005 Posts: 41 Location: United States Of America
|
|
|
|
I have a small question, If I want to list datasets starting with specific string is LVL work ?
Ex I have datasets like
MSA0455.FILE1
MSA0465.FILE2
MSA0488.FILE3
MSA0499.FILE4
In the above example, all datasets starting with "MSA04"
I tried with MSA04* in place of HLQ, but its failing. |
|
Back to top |
|
|
Douglas Wilder
Active User
Joined: 28 Nov 2006 Posts: 305 Location: Deerfield IL
|
|
|
|
Listcat uses whole nodes only. LMDLIST LISTID will use partial nodes. |
|
Back to top |
|
|
HappySrinu
Active User
Joined: 22 Jan 2008 Posts: 194 Location: India
|
|
|
|
Why don't you try contact your security team either ACF2/RACF team in your shop?
Who ever be the mainframe team, every one should contact them for the access of HLQ for certain teams.
More over they can run a report on certain machine for all HLQ's and owenrs also. I think I have asked that type report before.
Check with you security admins team. |
|
Back to top |
|
|
satya12
New User
Joined: 25 Apr 2006 Posts: 32 Location: India
|
|
|
|
Hi Anuj
I am working on mainframe apps decommission project.I came accross same situation at the beginning of the project.We have so many old apps sitting on mainframe no idea who is using them.I created a spread sheet with all of the high level qualifiers and sent it to RACF security adminstrator.These guys logs team name and apps steward information whenver they create a Profile.So i got the contact list for each qualifier and dealt with the apps from there.
Hope this helps you.
Thanks
Satya |
|
Back to top |
|
|
Anuj Dhawan
Superior Member
Joined: 22 Apr 2006 Posts: 6250 Location: Mumbai, India
|
|
|
|
Hi,
I don't have an easy access to these RACF guys, they can be contacted via proper (actually tedious) channel only. However, Thanks for sharing your experience, I'll try it, lets' see what's in store for me.. |
|
Back to top |
|
|
enrico-sorichetti
Superior Member
Joined: 14 Mar 2007 Posts: 10872 Location: italy
|
|
|
|
there are two maybe three issues discussed on in this thread....
which datasets belong to which teams and the relative security privileges
... ask the security support, they must know
SMF processing...
its a responsibility of the storage support for the sms records related to storage...
a very good tool to analyze smf record os the free:
DAF dataset Audit Facility, by Mke Cleary
File # 094 DAF from Mike Cleary - Dataset Audit Facility NEW
http://www.cbttape.org/ftp/cbt/CBT094.zip
but better search for the latest release
DCOLLECT processing
same as above for responsibilities
a reasonable good tool ( must be reviewed, but a very good starting point )
File # 206 DCOLLECT REXX execs from Linnea Nichols
http://www.cbttape.org/ftp/cbt/CBT206.zip
Remember, with the proper JCL DCOLLECT will collect also the relevant information
from the MCDS to give a proper picture of the datasets/volumes involved
these and lots of other useful stuff at
http://www.cbttape.org |
|
Back to top |
|
|
sameer
New User
Joined: 10 Mar 2005 Posts: 41 Location: United States Of America
|
|
|
|
Hello Douglas Wilder
Can you please elaborate how to use LMDLIST LISTID ? I basically want to display all datasets start with "MSA"
MSA* is not working when i tried to code LMDLIST in REXX program.
can't this be done using IDCAMS itself ? |
|
Back to top |
|
|
sameer
New User
Joined: 10 Mar 2005 Posts: 41 Location: United States Of America
|
|
|
|
Douglas Wilder,
Can you please help me how to use LMDLIST LISTID
Any sample code can help me better.
Thanks |
|
Back to top |
|
|
Anuj Dhawan
Superior Member
Joined: 22 Apr 2006 Posts: 6250 Location: Mumbai, India
|
|
|
|
Sameer,
Check this link
www.ibmmainframes.com/viewtopic.php?t=29466&highlight=lmdlist
might help.
Well, my thread is going in different direction...ah
Bill,
Yeah, instead of 'ALL' in LISTCAT 'NAME' gives me enough neat list, but I'm done with the list using IKJEFT1A. Main concern is
Quote: |
to find which DSN belongs to which team here at my shop? |
And what I thought, it happend the same way, I was suggested to do the following
Quote: |
create an excel of DSNs starting with that HLQ & send across the team to check which DSN belongs to which Team, then wait for their mail. |
and my sentence
Quote: |
Ah..it's a donkey's work. |
still stands..with grammer/spell correction.. , phew all of them are being lazy in updating the sheet even after one week.
does any one has some further suggestion ..? |
|
Back to top |
|
|
Anuj Dhawan
Superior Member
Joined: 22 Apr 2006 Posts: 6250 Location: Mumbai, India
|
|
|
|
Hi,
While working on above requirment LISTCAT gives me some datasets which are not on volume but are in system-catalog..
I mean, LISTCAT shows me these data sets when i execute an IDCAMS JOB but when I try to open them using ISPF 3.4, message is "No data set names found", what should I conclude ?
I think, one would say "DSNs are deleted but their catalog entry is not deleted", ritght? If yes, I would like to ass, why would one do this ? |
|
Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006 Posts: 19244 Location: Inside the Matrix
|
|
|
|
Hi Anuj,
Quote: |
I think, one would say "DSNs are deleted but their catalog entry is not deleted", ritght? If yes, I would like to ask, why would one do this ?
|
Usually, one does this out of ignorance - they do not know how create proper jcl. Sometimes people try to get too "cute" and create problems by implementing some "trick" they have learned.
Left to themselves, many a system gets out-of-sync. In the pre-sms days, many centers regularly ran jobs to identify catalog entries with no dataset and datasets with no catalog entry. These were then cleaned up. One of the old horror stories is the critical job that cannot allocate a particular output dataset. Seems like a file of that dsn was on every dasd volume . . . |
|
Back to top |
|
|
Anuj Dhawan
Superior Member
Joined: 22 Apr 2006 Posts: 6250 Location: Mumbai, India
|
|
|
|
Hi Dick,
Thanks for Your time...
Well, this is little new for me
Quote: |
Seems like a file of that dsn was on every dasd volume . . |
it's little tricky actually for me to understand, as mine is SMS-shop, this should haapen only on non-SMS site, perhaps I misunderstood something here, please assist. |
|
Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006 Posts: 19244 Location: Inside the Matrix
|
|
|
|
Hi Anuj,
Quote: |
this should haapen only on non-SMS site |
As i mentioned, this was in the pre-sms days.
It was quite common for badly written jcl to cause problems. Among the most common was propagating the same dataset across many dasd volumes. |
|
Back to top |
|
|
|