Maahjoor wrote:It's the user's work that would burn or use the CPU than the users. Even if it's some user, I doubt that all 300 users are burning the cpu to the max. And even further, it's fairly possible that nothing in the database is consuming the cpu but something running on the system is actually doing it. So if you are licensed to use AWR , either it's report or from the Statspack report, pull up a report and post it here.
i have windows server 2003, with oracle database 10g r2, 2.4 GHz (2 CPU) processor, and GB of RAM.
we have concurrently connected 300 users.
and it happens often that our connection to that database hangs, and no further connections are allowed to connect.
i check the performance tab in the WINDOWS TASK MANAGER, and i find the CPU 100% in use.Ram was 40%.
my oracle server is not dedicated server, it is shared server.
is this possible that 300 concurrent user could occopy usch a cpu?
there is a warning for dispatcher Warning: Total dispatchers (1) > MAX_DISPATCHERS (0).
Windows Server 2003 Version V5.2 Service Pack 2 CPU : 2 - type 586, 1 Physical Cores Process Affinity: 0x00000000 Memory (A/P) : PH:1162M/2038M, PG:4654M/5970M, VA:1953M/2047M Sun Feb 17 14:19:51 2013 Starting ORACLE instance (normal) LICENSE_MAX_SESSION = 0 LICENSE_SESSIONS_WARNING = 0 Picked latch-free SCN scheme 2 KCCDEBUG_LEVEL = 0 Using LOG_ARCHIVE_DEST_10 parameter default value as USE_DB_RECOVERY_FILE_DEST Autotune of undo retention is turned on. Dynamic strands is set to TRUE Running with 2 shared and 110 private strand(s). Zero-copy redo is FALSE IMODE=BR ILAT =110 LICENSE_MAX_USERS = 0 SYS auditing is disabled Warning: Total dispatchers (1) > MAX_DISPATCHERS (0) Starting up ORACLE RDBMS Version: 10.1.0.4.2. System parameters with non-default values: processes = 500 sessions = 1000 sga_max_size = 524288000 __shared_pool_size = 163577856 shared_pool_size = 150994944 __large_pool_size = 16777216 large_pool_size = 8388608 __java_pool_size = 67108864 java_pool_size = 67108864 sga_target = 524288000 control_files = E:\ORACLEAS\ORADATA\ORCL\CONTROL01.CTL, E:\ORACLEAS\ORADATA\ORCL\CONTROL02.CTL, E:\ORACLEAS\ORADATA\ORCL\CONTROL03.CTL db_block_size = 8192 __db_cache_size = 272629760 db_cache_size = 50331648 max_commit_propagation_delay= 0 compatible = 10.1.0.2.0 db_file_multiblock_read_count= 16 db_recovery_file_dest = E:\OracleAS\flash_recovery_area db_recovery_file_dest_size= 2147483648 undo_management = AUTO undo_tablespace = UNDOTBS remote_login_passwordfile= EXCLUSIVE db_domain = hct.org dispatchers = (protocol=TCP) shared_servers = 3 max_shared_servers = 3 max_dispatchers = 0 local_listener = vhost_alias job_queue_processes = 5 background_dump_dest = E:\ORACLEAS\ADMIN\ORCL\BDUMP user_dump_dest = E:\ORACLEAS\ADMIN\ORCL\UDUMP core_dump_dest = E:\ORACLEAS\ADMIN\ORCL\CDUMP db_name = orcl open_cursors = 300 pga_aggregate_target = 33554432 aq_tm_processes = 1