Conference utrop1::lnx_apo

474.0. "Field om_id in table objekt - what if two data cells?" by KETJE::HAENTJENS (Beware of Counterfeit) Wed Apr 30 1997 11:07

    If a second data cell is installed on a server, it requires a separate
    database, e.g. celldb2, but it will use the same pool root directory.
    This is what I read in the system management guide.
    
    The database table "objekt" in the cell database has the basic
    information about objects. Does it then only store information on
    objects of a single data cell? If yes, then why does it have a field
    om_id? It would always have the same value...
    
    I'm not asking this question only to be annoying. We have a program
    that reads the table "objekt" and I'm trying to prepare the ground for
    the customer's next step. He wants to get rid of one LNX server, so we
    have to move the cell onto another server as second data cell.
    
    The program just reads the table "objekt" and it exists because with
    the current limitations on APO/Query we could not figure out another
    clever way to run over all objects.
    
    Rene Haentjens.

Why don't you use 'dbu' to backup the cell and restore it
on the other system?

Jos

Or is your concern this path in the pool?
The pool root dir will be different for each cell

/usr/op/cell/pool/00000001/..  for the first cell
/usr/op/cell/pool/00000002/..  for the second..

So that's not a problem.

Regards,
Jos

    I should have been more precise in what I want to know.
    
    If my program runs over all entries in the table "objekt", will it then
    find all objects of a single data cell and nothing else? Or will it also 
    find other objects? If yes, which ones?
    
    Ren�.