"Natra" <natra_at_libbbero.it> wrote in message
news:ariioh$c0f$1_at_newsreader.mailgate.org...
> Hi boys,
> I've a great problem on a Linux Box Red Hat 7.2 with Oracle 817 with patch
> (hardware: P3 1000 Mhz - 512 MbRam).
> Before installing the Linux Box mi database worked on a Ora 816 (on the
same
> hardware) on WinNT (SerPack 6a).
> In order to migrate DB to the new OS I've used the export utility from the
> command prompt.
> After the installation of new enviroment I've operate the Import ( from
> Linux shell command line) of .dmp file (160 Mb) and all seems to be done
> very well.
> After installing the ERP procedure that uses Oracle DB I've immediately
> verified an incredible slowly performance in quite simple query (with
order
> by) in a 50.000 row table, (this query in NT environment takes arount 50
> sec. with a 99% CPU time - in the new environment takes around 28 minutes
> with 98 % of CPU time ad around 14% of memory). Other that situation we
have
> verified that the whole operation in DB was greatly decrease after import.
> I've tried to modify a lots of time the init file but the increase of
> performance was imperceptible.
> I've read in some posts that it's possible that the recreation of the DB
> dictionary indexes during import was failed ant that must ricreate ( or
> optimizate ) with some ORACLE command, that I don't know.
> If exist some command to help me in this situation, please post it.
> Thanks for attention.
>
> Francesco Ortolani
>
> P.S.
> I'm sorry for my english... I'm italian :-)
>
>
> --------------------------------------------------------------------------
--
> ----------
> This is my actual init file
> --------------------------------------------------------------------------
--
> ----------
>
> db_name = "colfi"
> instance_name = colfi
> service_names = colfi
>
> control_files = ("/home/oracle/oradata/colfi/control01.ctl",
> "/home/oracle/oradata/colfi/control02.ctl",
> "/home/oracle/oradata/colfi/control03.ctl")
>
> db_files=1024
>
> open_cursors = 300
> #open_cursors = 256
>
> max_enabled_roles = 30
> db_file_multiblock_read_count = 32
> db_block_buffers = 45000
>
> shared_pool_size = 48321843
> # shared_pool_size = 9000000
>
> large_pool_size = 614400
> java_pool_size = 0
>
> log_checkpoint_interval = 10000
> log_checkpoint_timeout = 1800
>
> processes = 150
> #processes = 100
>
> log_buffer = 163840
> #log_buffer = 32768
>
> parallel_max_servers = 5
> max_dump_file_size = 10240
>
> rollback_segments = ( rbs0, rbs1, rbs2, rbs3, rbs4, rbs5, rbs6, rbs7,
rbs8,
> rbs9, rbs10, rbs11, rbs12, rbs13, rbs14, rbs15 )
>
> background_dump_dest = /home/oracle/admin/colfi/bdump
> core_dump_dest = /home/oracle/admin/colfi/cdump
> resource_manager_plan = system_plan
> user_dump_dest = /home/oracle/admin/colfi/udump
>
> db_block_size = 8192
>
> remote_login_passwordfile = exclusive
>
> os_authent_prefix = ""
>
> job_queue_processes = 2
> #job_queue_processes =1
>
> job_queue_interval = 30
> distributed_transactions = 500
> open_links = 4
>
> mts_dispatchers = "(PROTOCOL=TCP)(SER=MODOSE)"
> mts_dispatchers = "(PROTOCOL=TCP)(PRE=oracle.aurora.server.SGiopServer)"
>
> dml_locks = 748
> compatible = "8.1.0"
> sort_area_size = 65536
>
>
>
I haved the same problem. It was wrong "importfrom.dmp" file. Check it.
Received on Thu Nov 21 2002 - 07:10:55 CST
