sql - Populating table B with data from table A if data in table B is different to table A -

apolgies confusing heading , long winded post.

i attempting populate (oracle) sql table data can monitor on regular basis , build picture of users use application modules.

i have table 'forms_sessions_log' created follows:

create table stuman.forms_sessions_log ( select a.sid, a.serial#, a.logon_time, a.client_identifier, a.module,a.username, a.schemaname, b.spriden_last_name, b.spriden_first_name gv$session a, spriden b program 'frmweb%' , a.username=b.spriden_id) 

this creates 'snapshot' of 'forms' sessions connected database @ point in time.

from on, each minute via job scheduler, i'd continue populate table new or different sessions.

i came following:

insert stuman.forms_sessions_log (sid,serial#,logon_time,client_identifier,module,username,schemaname,spriden_last_name,spriden_first_name) select  a.sid,          a.serial#,          a.logon_time,          a.client_identifier,          a.module ,          a.username,          a.schemaname,          b.spriden_last_name,          b.spriden_first_name    gv$session          spriden b       a.program 'frmweb%' ,         a.username=b.spriden_id          ,         not exists(select * stuman.forms_sessions_log) / 

this didn't seem work. failed add new sessions. if truncate stuman.forms_sessions_log , run insert statement again, populates fine fails add new sessions table on consecutive attempts.

i looked @ code , decided fine tune statement using comparisons in not exists clause.

insert stuman.forms_sessions_log   (sid,serial#,logon_time,client_identifier,module,username,schemaname,spriden_last_name,spriden_first_name) select  a.sid,      a.serial#,      a.logon_time,      a.client_identifier,      a.module ,      a.username,      a.schemaname,      b.spriden_last_name,      b.spriden_first_name    gv$session a,          spriden b       a.program 'frmweb%' ,         a.username=b.spriden_id          ,         not exists(select     a.sid,a.serial#,a.logon_time,a.client_identifier,a.module,a.username,a.schemaname              stuman.forms_sessions_log a,                   gv$session b           a.sid=b.sid          ,     a.serial#=b.serial#         ,     a.logon_time=b.logon_time         ,     a.client_identifier=b.client_identifier         ,     a.module=b.module          ,     a.username=b.username         ,     a.schemaname=b.schemaname                 ) / 

this doesn't expecting. isolated not exists part of query wanted check is:

select     a.sid,a.serial#,a.logon_time,a.client_identifier,a.module,a.username,a.schemaname              stuman.forms_sessions_log a,                   gv$session b           a.sid=b.sid          ,     a.serial#=b.serial#         ,     a.logon_time=b.logon_time         ,     a.client_identifier=b.client_identifier         ,     a.module=b.module          ,     a.username=b.username         ,     a.schemaname=b.schemaname 

now, bear in mind defined fields in both table , b identical (stuman.forms_session_log populated gv$session), nothing returns.

i have 1 row stuman.forms_session_log corresponds row in gv$session.

why select statement above not returning results?

have tried user merge?

merge stuman.forms_sessions_log sl using (select  a.sid,                 a.serial#,                 a.logon_time,                 a.client_identifier,                 a.module ,                 a.username,                 a.schemaname,                 b.spriden_last_name,                 b.spriden_first_name           gv$session a,                 spriden b          a.program 'frmweb%'         ,    a.username=b.spriden_id  ) s on (    s.sid = sl.sid      , s.serial# = sl.serial#) when not matched insert (sid,serial#,logon_time,client_identifier,module,username,schemaname,spriden_last_name,spriden_first_name)   values (s.*)