ADSM-L

Re: Minimum requirements for ADSM server recovery

1997-06-03 15:58:53
Subject: Re: Minimum requirements for ADSM server recovery
From: Tim Dobrowolsky <tim AT CATRS.CAT.CC.MD DOT US>
Date: Tue, 3 Jun 1997 15:58:53 -0400
In case you are wondering, this is for an alternate offsite backup, last
ditch thing.
I'm trying to minimize the amount of human work and tapes etc that are
needed daily.
Painful work is acceptable to perform the recovery - as long as it CAN be
performed.
The database space eaten by the copy pool is okay.

Thanks for your response.


At 01:59 PM 6/3/97 -0500, you wrote:
>     Even though our AIX ADSM environments were our first to be created, we
>     (I) have only tested recovery of our MVS environment(s).
>     Now our diskpool is dumped to a tapepool that is itself offsite so we
>     don't use copypool (thank god 'cause they beat the heck out of our
>     servers)
>     BUT
>     you'll find these things helpful...
>     (were the different server environments naturally have different
>     requirements in general, some things MIGHT be excessive but work !)
>     1) a copy of your device configuration file ! (so you can read your
>        db backup tape)

    I expect I can reconfigure by hand, yes?  I will have prints of this info.

>     2) a "q dbvol f=d"
>     3) a "q logvol f=d"
>     4) a "q volume dev=disk"
>     5) a 'as current as can be had' copy of your volume history file
>        OR a "q volhist" OR BOTH !
>        I keep a copy of the volhist for layout examples and then every
>        3 hours on a remote box run a "q volhist begind=today begint=00:00"
>        you can always type entries in by hand but you can't pull the info
>        from thin air !
>
    I take it you need to have a entry in the vol history in order to use it?
    For instance, if I have a volume "DB4" which contains a db backup, I can't
    just use it in the standalone recovery?  Since the backup will be stored
with
    the storage pool it will be evident from the label which it volume it is.

>     *Make replacement db & log files the same size as origionals

    Makes sense.

>     *Start small & build  ie. create your 1st db & log files as on your
>        origional environment & start adsm... diskpool, we don't need no
>        stinking diskpool

    Not at first, anyway - OK

>     *With a running ADSM, add additional DB & LOG files to match the lost
>        environment
>     *RESTORE THE DB
>     *Do voodoo magic to DELETE EXPECTED TO EXIST DISKPOOL VOLUMES !
>        (ie see the book on your server platform)
>     *Start ADSM and now worry about a diskpool, auditing seq. vols., etc.
>
>     GOOD LUCK
>
>     it really isn't that bad.
>
>     later
>          Dwight
>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
> Tim Dobrowolsky                   > <                yksloworboD miT <
> Computer Services                 > <              secivreS retupmoC <
> Catonsville Community College     > <  egelloC ytinummoC ellivsnotaC <
> tim AT catrs.cat.cc.md DOT us            > <         su.dm.cc.tac.srtac@mit <
> 410-455-4562                      > <                   2654-554-014 <
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
<Prev in Thread] Current Thread [Next in Thread>