1 Reply Latest reply on Apr 2, 2013 10:56 PM by clcarter

    XE 11.2  without using apex (& other questions)

      I've created an XE instance to take the place of a smaller 10G database currently running on EE we have, because for the size of it and use it gets XE is more than adequate.

      I find the XE interface rather limited (compared to original XE) so I'm assuming most things like creating users, tablespaces etc will need to be done with sqlplus.

      I also see that apex is part of it now, and I don't want to use apex (unless I have to)

      I manually created the tablespace and three users from the 10G database I want to bring across.

      Here are my questions -

      for getting the 10G data across, should I just use impdp to get the schema(s) I want? Only one of the above users has any tables or other objects, or should I use full = y impdp?

      I did a full=y impdb as a trial run, I see it brought over sysman, mgmt_view etc.. (as well as many error messages) - or are these default XE users as well? if not, can or should I just drop them?

      I just want an existing application to connect to the database with the user account I created - I don't need to create a workspace or anything in apex for it do I?

      thanks in advance

      Edited by: user571263 on Apr 2, 2013 3:25 PM
        • 1. Re: XE 11.2  without using apex (& other questions)
          I don't want to use apex
          So don't use it. Disable it if you'd like to be be sure the web pages won't be accessed, quickest way to whack apex is remove the shared server parameters.
          sqlplus /nolog
          connect system
           ... password ... connected. 
          alter system reset dispatchers scope=spfile;
          alter system reset shared_servers scope=spfile;
          -- and bounce to enforce changes and make sure spfile is good
          conn /as sysdba;
          shutdown immediate;
          I'm pulling those out from memory, one/the other/or both might be 'alter database ...'. Try 'em, if it gripes, try ' ... database'.
          should I just use impdp ... get the schema(s) I want?
          That's one of the recommended ways, schema export/import. The full 11g documentation, the Utilities book, has the details. If tablespace(s) and/or schema name(s) need adjustment on the impdp end, or exclude some objects, or just include a subset of objects, there are options to specify what goes where and what gets imported.


          I never use exp/imp or datapump full=y for anything, it can muck just about anything in a database instance outside of the sys/system catalogs if not done with care. One or three schemas at a time is a good approach, although I've seen folks trying to expdp over 1,000 schemas in one pass. Yuck. I'd use a tablespace transport for that chore. Doesn't take much longer than copying the datafiles over the network, if the source and target OS both use the same endian-ness CPU.
          sysman ...mgmt_view ... errors
          Those are schemas and roles used by the oracle agent and dbsnmp schema for instance health monitoring, maybe no harm done if you're not using the agent or don't have an OMS. I suppose those schemas could be dropped, but I think sysman could have some memory or other internal monitoring/health check/tuning chores, MMAN, MMON processes, perhaps? I'm not positive.
          application to connect to the database ... create a workspace ...?
          Nope, if you're not using apex, in that case workspaces are not relevant.