當(dāng)然，你可以找到oracle自帶的init模板，一個個參數(shù)地設(shè)置自己系統(tǒng)的參數(shù)文件。那這將是一件很糟的事情，它會浪費你寶貴的時間。

解決方法：

從alert_alaska.log警告日志里著手，因為它記錄著一直以來數(shù)據(jù)庫運行的情況，當(dāng)然也包括每次啟動的參數(shù)信息啦，我們要的就是:

processes = 150

__shared_pool_size = 75497472

__large_pool_size = 4194304

__java_pool_size = 4194304

__streams_pool_size = 0

nls_language = AMERICAN

nls_territory = AMERICA

sga_target = 167772160

control_files = E:/ORACLE/PRODUCT/10.2.0/ORADATA/ALASKA/CONTROL01.CTL, E:/ORACLE/PRODUCT/10.2.0/ORADATA/ALASKA/CONTROL02.CTL, E:/ORACLE/PRODUCT/10.2.0/ORADATA/ALASKA/CONTROL03.CTL

db_block_size = 8192

__db_cache_size = 79691776

compatible = 10.2.0.1.0

db_file_multiblock_read_count= 16

db_recovery_file_dest = e:/oracle/product/10.2.0/alash_recovery_area

db_recovery_file_dest_size= 1073741824

log_checkpoints_to_alert = TRUE

undo_management = AUTO

undo_tablespace = UNDOTBS1

undo_retention = 900

remote_login_passWordfile= EXCLUSIVE

db_domain = com.cn

dispatchers = '(PROTOCOL=TCP) (SERVICE=alaskaXDB)'

#用 ' ' 引起它們

job_queue_processes = 10

audit_file_dest = E:/ORACLE/PRODUCT/10.2.0/ADMIN/ALASKA/ADUMP

background_dump_dest = E:/ORACLE/PRODUCT/10.2.0/ADMIN/ALASKA/BDUMP

user_dump_dest = E:/ORACLEPRODUCT/10.2.0ADMIN/ALASKA/UDUMP

core_dump_dest = E:/ORACLE/PRODUCT/10.2.0/ADMIN/ALASKA/CDUMP

db_name = alaska

open_cursors = 300

pga_aggregate_target = 16777216

復(fù)制以前成功啟動的參數(shù)語句(如上代碼段)到一個文本中，只需把dispatchers = (PROTOCOL=TCP) (SERVICE=alaskaXDB)的值加上''（單引號），變成dispatchers = '(PROTOCOL=TCP) (SERVICE=alaskaXDB)'，保存為c:/pfile.txt

最后運行：startup pfile='c:/pfile.txt';即可完成。