Tom,

We have also made this discovery and are none too pleased about it.  We
have an open PMR on this, but I don't think I've made much headway in
convincing support/development that this is a problem.  As I understand it,
this change was made to address some design change requests (DCRs), having
to do with being able to restore a full directory structure using archives,
just as you can do with backups.  I understand the desire to satisfy the
DCR, but the side effect of the implementation has very adversely affected
our existing use of archives, in the way that you described (below).

As with your example, we also use ADSM archive to archive hundreds or
thousands of files from the same directory.  With the new archive
behaviour, this will result in tons of directory turds getting left in the
ADSM DB after we remove the files - if I understand the process correctly.
This simply does not work for us.

..Paul
--
At 07:04 AM 3/10/99 -0500, Fluker, Tom R wrote:
At 07:04 AM 3/10/99 -0500, Fluker, Tom R wrote:
>It appears to me that, when I archive a file, a database entry gets created
>for every directory 'above' the file I'm archiving.
>
>For example, if I "dsmc archive /usr1/mygroup/myhome/myfile1" then a
>database entry would get created for /usr1/mygroup, /usr1/mygroup/myhome,
>and /usr1/mygroup/myhome/myfile1.  This is, perhaps, understandable.
>
>If I then "dsmc archive /usr1/mygroup/myhome/myfile2" a second database
>entry for both /usr1/mygroup and /usr1/mygroup/myhome would get created.
>
>If I archive 1000 files from the same directory, three levels deep, I
>(apparently) get an additional 2000 database entries that don't seem
>necessary.
>
>Is this true?  If so, is there anyway to eliminate the duplicate database
>entries for the same directory levels.  As I understand it ADSM doesn't
>"clean up" these directory entries from the database when file file is
>removed anyway.  Is that true?  Ugh!!!
>
>Tom Fluker
>tom.fluker AT viasystems DOT com
>