Hi,
We are using version 4.5 of OTM and have 10 or so basic queries that populate users Business Monitors. Under normal circumstances each query takes 10-30 secs to execute but occasionally one of these queries will lock up the database server taking close to 100% CPU until the SQL session is killed (and the system becomes unusable until this happens). The client has a very short timeout on their HTTP proxy (2 minutes) and the problem seems to be related to when a timeout occurs during population of the BM (or during Finder set searches). Oracle suggest that the queries are too complex and should be simplified - one recent example query was 'select count(*) from ob_order_base'. I am still awaiting a response to exactly how this query can be simplified!
Has anyone seen anything similar or have any ideas as to what the issue might be.
Any thoughts greatly appreciated. Solutions or explanation even more welcome.
We are using version 4.5 of OTM and have 10 or so basic queries that populate users Business Monitors. Under normal circumstances each query takes 10-30 secs to execute but occasionally one of these queries will lock up the database server taking close to 100% CPU until the SQL session is killed (and the system becomes unusable until this happens). The client has a very short timeout on their HTTP proxy (2 minutes) and the problem seems to be related to when a timeout occurs during population of the BM (or during Finder set searches). Oracle suggest that the queries are too complex and should be simplified - one recent example query was 'select count(*) from ob_order_base'. I am still awaiting a response to exactly how this query can be simplified!
Has anyone seen anything similar or have any ideas as to what the issue might be.
Any thoughts greatly appreciated. Solutions or explanation even more welcome.
Comment