Long response times calling XE on a Linux Server

Question

I've installed XE on a Suse 9.2 Linux Server.
In most cases I get a fast access to XE over APEX. But sometimes, I get a timeout when a I call a page on APEX.
There is also a paralell running Apache webserver (port 80) on the server.
XE port 8080.
Does anybody know what's the reason for this?

Answer 1

you may catch cpu consuming processes from operating system and if they are related to xe instance you can use this easily to identify what is going on behind;
-- who is doing what
SELECT sq.sql_text,
se.status,
se.username,
se.osuser,
se.terminal,
se.machine,
'ALTER SYSTEM KILL SESSION ''' || se.sid || ', ' || se.serial# ||
''' IMMEDIATE ;',
se.sid,
se.serial#,
pr.spid,
se.saddr,
se.schemaname,
se.program,
se.logon_time,
sq.hash_value,
sq.address,
sq.buffer_gets,
sq.disk_reads,
sq.executions,
sq.fetches,
sq.rows_processed,
io.block_gets,
io.consistent_gets,
io.physical_reads,
io.block_changes,
io.consistent_changes,
se.module,
se.action,
se.client_info,
sw.event,
sw.p1text,
sw.p1,
sw.p2text,
sw.p2,
pr.pga_used_mem,
pr.pga_alloc_mem,
pr.pga_freeable_mem,
pr.pga_max_mem,
'exec dbms_support.start_trace_in_session(' || se.sid || ', ' || se.serial# || ', ' ||
'waits => TRUE' || ', ' || 'binds => TRUE);',
'exec dbms_support.stop_trace_in_session(' || se.sid || ', ' || se.serial# || ');'
FROM v$session se, v$sqlarea sq, v$process pr, v$session_wait sw, v$sess_io io
WHERE status IN ('ACTIVE', 'KILLED', ' INACTIVE', ' CACHED', ' SNIPED')
AND se.sql_hash_value = sq.hash_value(+)
AND se.sql_address = sq.address(+)
AND se.paddr = pr.addr
AND se.sid = sw.sid
AND se.sid = io.sid
ORDER BY sq.sql_text, se.username, se.schemaname, se.osuser, se.program, se.sid
You need to match the top cpu os processes' id with the "pr.spid" column, then you might look at the other columns of that row for details.
Also the database runs a number of jobs - backups, etc. these may be the problems.
Kind regards,
Tonguç