How do YOU keep track of what primary stgpools _ought_ to be backed up? Not what are we doing, but what should we be doing?
There's a set of data that we all need to keep track of, which TSM simply punts: 'where should this primary stgpool be backed up to?'. I don't know about most of you, but I have stylized stgpool names, which let me identify candidate copy pools from nomenclature alone. YADDA-DISK YADDA-TAPE YADDA-OPTICAL all back up to YADDA-C1 and YADDA-C2 and so forth, if they are there. But this is an immensely clumsy way to do it. I want a way to set, in the TSM database, the notion "BLOTZMO-DISK, WAKKO-TAPE, and AAGRAVAAG-OPTICAL should be backed up to ARCHIPELAGO-MOOSETHROTTLE". Then I could issue a command of the sort: back stgpool BLOTZMO-DISK figure_it_out_youre_the_computer=yes Not only does the lack of recordkeeping make it impossible to do automatic maintainance as I suggest, it makes it a priori impossible to write a monitoring tool without maintaining some sort of external configuration file to hold the data. This could be solved officially from the inside, or it could be solved by letting us add "columns" to the "table" STGPOOLS. We could come up with standardized names for this kind of data, and we'd at least be able to do -something- with the base package. Heck, even a single-value field BACKSTGPOOL to match NEXTSTGPOOL would be something. Of course, to do it right you'd want a relation: there's no technical reason to prevent peole from backing up all their primary stgpools to all their copypools. - Allen S. Rout - NERDC TSM admin
