Stan reminds us...


> Yes, I agree.  Before jumbo datasets were released, I suggested (to
> all backup product engineers) that if they ever see a root file in
> a backup list (no matter how the list was specified), that they should
> default to thinking "oh, the user probably wants the entire database
> backed up" ... and that they should add a keyword like "NODB" (or
> something)
> to *defeat* that default.
>
> In fact, if you ask for statistics on the Turbostore command line, you
> will be told about  missing datasets (taken I presume, from the root
> file), and will have a section detailing the backup (successful or
> otherwise) of entire databases. Another fine example of the close
> integration of H-P's native tools in MPE.
>
>
> First, before the store starts
>
>  /DISC/PUB/EMAST01 NOT STORED: FILE NOT FOUND
>  /DISC/PUB/EMAST02 NOT STORED: FILE NOT FOUND
>  /DISC/PUB/EMAST03 NOT STORED: FILE NOT FOUND
>  /DISC/PUB/EMAST04 NOT STORED: FILE NOT FOUND
>
> Then, when it's done
>
>  TURBOIMAGE DATABASE: /DISC/PUB/EMAST
>    FILES STORED :                                        1
>    FILES FAILED TO STORE :                               4
>
>    TURBOIMAGE DATABASE: /DISCTRL/DEMODB/ACCT
>    FILES STORED :                                       15
>
>    TURBOIMAGE DATABASE: /DISCTRL/DEMODB/OMNIDD
>    FILES STORED :                                       14
>
>    TURBOIMAGE DATABASE: /DISCTRL/DEMODB/OSTORE
>    FILES STORED :                                        6
>
> Neil Harvey
>