Looking into trace file that must be written because of the core dump would
be of more importance I guess.
Check the errors that occured on Metalink.
Epicentre team A <oracle_at_epicentre.fr> schreef in berichtnieuws
ak7jj9$n9e$1_at_wanadoo.fr...
| Hi
|
| I want to change parameter file but when massives imports via Sql Loader
| import stop about 15 minutes. There is importants core dump errors.
| On server no problem with pagging or swapping operations (normal)
|
| Platform: HP-UX
| Oracle Enterprise 8.1.7
| Ram 2 Go
| Raid 1
|
|
|
| Before change we have.
|
| db_files=200
| large_pool_size=614400
| shared_pool_size=80485760
| hash_multiblock_io_count=0
| db_block_lru_latches=2
| cursor_sharing_force=exact
| optimizer_index_caching=50
|
| After Change it's
|
|
| db_files=150
| large_pool_size=300000
| shared_pool_size=81920000
| hash_multiblock_io_count=8
| db_block_lru_latches=4
| cursor_sharing_force=force
| optimizer_index_caching=99
|
|
| I've you an idea ?
| Denis Jeanneret
|
|
Received on Sat Aug 24 2002 - 08:54:55 CDT
