Getting quite a number of zombie processes in MySQL from LAMS. Seems to be a failed query (included below). Each process hogs 10% of CPU, so after a few days the server sits at 70% CPU (64 bit Win 2008 Server)
select distinct event0_.uid as uid0_, subscripti1_.uid as uid1_, event0_.default_message as default2_43_0_, event0_.default_subject as default3_43_0_, event0_.name as name43_0_, event0_.scope as scope43_0_, event0_.event_session_id as event6_43_0_, event0_.triggered as triggered43_0_, event0_.fail_time as fail8_43_0_, event0_.message as message43_0_, event0_.subject as subject43_0_, subscripti1_.last_operation_message as last2_44_1_, subscripti1_.periodicity as periodic3_44_1_, subscripti1_.user_id as user4_44_1_, subscripti1_.delivery_method_id as delivery5_44_1_, subscripti1_.last_operation_time as last6_44_1_, subscripti1_.event_uid as event7___, subscripti1_.uid as uid__ from lams_events event0_ left outer join lams_event_subscriptions subscripti1_ on event0_.uid=subscripti1_.event_uid, lams_event_subscriptions subscripti2_, lams_event_subscriptions subscripti3_, lams_event_subscriptions subscripti4_ where event0_.uid=subscripti4_.event_uid and event0_.uid=subscripti3_.event_uid and event0_.uid=subscripti2_.event_uid and (event0_.fail_time is not null or subscripti2_.periodicity>0 and now()-subscripti3_.last_operation_time>=subscripti4_.periodicity) order by subscripti1_.last_operation_time desc
Posted by Peter Vitez
|
