Delete Multiple PDS

Santty · New User Joined: 20 Nov 2007 Posts: 37 Location: USA

Hi,
I need to delete N number of dataset. They are of pattern ABC.DEF.GHI.*

But i have some datasets as below
ABC.DEF.GHI.D1
ABC.DEF.GHI.D2
ABC.DEF.GHI.D1.T1
ABC.DEF.GHI.D1.T2
ABC.DEF.GHI.D2.T1
ABC.DEF.GHI.D2.T2
ABC.DEF.GHI.D2.T3

Need to delete nearly 50k datasets in this combination (ABC.DEF.GHI.*)

If i use delete in IDCAMS

dick scherrer · Posted: Mon Oct 29, 2012 11:30 pm

Hello,

You might look into using ADRDSSU/DFDSS. . .

enrico-sorichetti · Posted: Mon Oct 29, 2012 11:54 pm

see here for a discussion of ADRDSSU filtering

www.ibmmainframes.com/viewtopic.php?t=29913&highlight=adrdssu

Santty · New User Joined: 20 Nov 2007 Posts: 37 Location: USA

Thank you Dick scherrer & Enrico-sorichetti

ADRDSSU works in the same way as IDCAMS

enrico-sorichetti · Posted: Tue Oct 30, 2012 12:35 am

Santty · New User Joined: 20 Nov 2007 Posts: 37 Location: USA

Thanks a lot Enrico-sorichetti

It works ....

enrico-sorichetti · Posted: Tue Oct 30, 2012 12:51 am

I hope that You tested with a TYPRUN=NORUN
before going live

and read about the BY clause to find out about other filtering constructs

Santty · New User Joined: 20 Nov 2007 Posts: 37 Location: USA

No Enrico ... i tested it with the parm 'TYPRUN=NORUN' before going live ...

ibmmainframes.com/about6069.html

dick scherrer · Posted: Tue Oct 30, 2012 1:08 am

Hello,

Also (probably way too late for this exercise), in addition to running a NORUN, it is often desirable to create a backup of the data before deleting it. How ugly might things become if someone says Oops - we really need to get some of that back?

Santty · New User Joined: 20 Nov 2007 Posts: 37 Location: USA

Yes, you are correct ....

gcicchet · Posted: Tue Oct 30, 2012 3:32 am

Hi,

search on both IDCAMS and MASK.

It may also assist you.

Gerry

mmwife · Super Moderator Joined: 30 May 2003 Posts: 1592

Hi Santty,

Why don't you give .*.* a shot?

gcicchet · Posted: Sat Nov 03, 2012 2:26 pm

Hi Jack,

bad shot

It doesn't work.

Gerry

mmwife · Super Moderator Joined: 30 May 2003 Posts: 1592

Still, worth a shot.

Pete Wilson · Posted: Mon Nov 05, 2012 9:30 pm

You can use the IDCAMS DELETE MASK parameter with the % and/or * masks to delete generic dataset masks if you're at about ZOS1.11 or above.

Look at the IDCAMS manual in the DELETE command and it's MASK parameter

Pete Wilson · Posted: Mon Nov 05, 2012 9:33 pm

In z/OS V1R11, DFSMS access method services (IDCAMS) adds a new MASK option to the DELETE command. This option lets you specify many variations of a data set name on a single deletion, using new wild card characters and rules to give more flexibility in selecting the data sets to be deleted.

|Previously, only one data set qualifier could be |replaced by a wild card on a DELETE command. For example: DELETE A.*.C |would delete data sets with A as the first qualifier, any second qualifier, |and C as the third qualifier, such as A.B.C or A.BB.C. With the new |MASK keyword, you can replace multiple qualifiers with wild cards, |and can replace specific characters in a qualifier name with wild |cards as well.

|The MASK keyword treats a single asterisk in the |same way as the generic DELETE command. For example DELETE A.*.C MASK |specifies a data set with A and C as the first and third qualifiers, |and a second qualifier with any set of characters. In addition, the MASK keyword allows two consecutive asterisks to |replace multiple qualifier names. For example, the entry name DELETE |A.**.B MASK means all data set names of two or more levels with A |as the first qualifier, and B as the last qualifier. A.B.** means |all data set names of two or more levels where A is the first qualifier |and B is the second (or last) qualifier. Double asterisks cannot be |part of a qualifier name; they must be preceded or followed by either |a period or a blank character.

|The MASK keyword also accepts percent signs (%) was |wild cards to replace individual characters in the same position. |You can specify from one to eight percent signs in each qualifier. | For example, DELETE A%DE MASK would match a data set named ABDE, |while DELETE A%%D% MASK would match one named ABCDE.

The DELETE MASK command allows only one data set entry-name to be specified. If multiple entry-names are specified, the DELETE request will fail with error messages. If more than 100 data set names are filtered from the wild card notation, AMS only deletes the first 100 data sets identified by the filtering process.

If you do not specify the MASK keyword on the DELETE command, or explicitly specify NOMASK, the previous wild card rules remain in effect (one asterisk can replace one qualifier in a data set name).