Log file switch (archiving needed)

Question

Hi,
My database is on windows 2003, 10.2.0.4, recently I have been getting the following wait events on a regular basis, LOG FILE SWITCH (ARCHIVING NEEDED)...My redo files are 50MB with 5 groups...I have changed the log_archive_max_process from 2 to 3, but still the problem persists...can anyone help me regarding the same?As to what other changes can be done?and this happens only during the time of BOD and EOD.
Thanks,
AJ

Answer 1

Hi Jonathan,
Thanks for the reply, I am not worried abt the archiving thing for the time being,as my database was working absolutely fine before the new query that was added in the Begin of Day process. So I am trying to tweak the query.
Folllowing is the procedure which gets called during the Begin of Day:
CREATE OR REPLACE procedure Lms_Pr_Bod_Update_Tmp (p_import_type in number)
as
v_update_script varchar2(2000);
v_upper_start_tag varchar2(10); -- san_22-apr-2009
v_upper_end_tag varchar2(10); -- san_22-apr-2009
begin
declare
cursor cur_update_tmp is
select -- this query is for all of lov type of data
a.destination_column_name, a.column_to_update , b.mapped_lov_syscode lov_syscode,
c.destination_table_name,'LOV_DATA_SYSCODE' select_attribute,'LMS_LOV_DATA_MAP' select_table,
'SOURCE_UNIQUE_ID' where_condition,'LOV_SYSCODE' lov_condition,'LOV' att_type
from lms_import_column_info a
inner join lms_attribute_master b
on a.attribute_syscode = b.attribute_syscode
and (a.column_to_update is not null or trim(a.column_to_update) = '')
inner join lms_import_type_master c
on a.import_type_syscode = c.import_type_syscode
where a.import_type_syscode = p_import_type
and b.mapped_lov_syscode is not null
union all
select -- this query is for all of dummy attribute
a.destination_column_name, a.column_to_update , 1 lov_syscode,
d.destination_table_name,c.attribute_name_internal select_attribute,
case
when b.applicable_for = 'INS' then 'LMS_ENTITY_INSTRUMENT'
when b.applicable_for = 'ACC' then 'LMS_ENTITY_ACCOUNT'
end case,
b.attribute_name_internal where_condition,'1' lov_condition,'DUMMY' att_type
from lms_import_column_info a
inner join lms_attribute_master b
on a.attribute_syscode = b.attribute_syscode
and (a.column_to_update is not null and trim(a.column_to_update) <> ' ')
and b.attribute_type = 'DUM' and dummy_column_type = 'FILT'
inner join lms_attribute_master c
on b.dummy_mapped_attribute_pk = c.attribute_syscode
inner join lms_import_type_master d
on a.import_type_syscode = d.import_type_syscode
where a.import_type_syscode = p_import_type;
begin
v_upper_start_tag := 'UPPER('; -- san_22-apr-2009
v_upper_end_tag := ')';          -- san_22-apr-2009
for cr_update_tmp in cur_update_tmp loop
if p_import_type = 4 and cr_update_tmp.column_to_update IN ('INSTRUMENT_SYSCODE','ACCOUNT_SYSCODE')
and Upper(cr_update_tmp.destination_table_name) = 'LMS_ENTITY_TRANSACTION_TMP' then
v_upper_start_tag := ''; -- san_22-apr-2009
v_upper_end_tag := ''; -- san_22-apr-2009
end if;
v_update_script := 'UPDATE ' || cr_update_tmp.destination_table_name || ' A SET ' || ' A.' || cr_update_tmp.column_to_update || ' =
(select ' || CR_UPDATE_TMP.SELECT_ATTRIBUTE || ' from ' || CR_UPDATE_TMP.SELECT_TABLE || '
where ' || V_UPPER_START_TAG || CR_UPDATE_TMP.WHERE_CONDITION || V_UPPER_END_TAG || '=' || V_UPPER_START_TAG || CR_UPDATE_TMP.DESTINATION_COLUMN_NAME || V_UPPER_END_TAG ||
' AND ' || cr_update_tmp.lov_condition || ' = ' || cr_update_tmp.lov_syscode || ')';
v_upper_start_tag := 'UPPER('; v_upper_end_tag := ')';
execute immediate (v_update_script);
end loop;
end;
end;
Following is the flow of the query:
1. A temporary table is created where in the updates can be made.
2. Rows are inserted into ths table from the source table
3. Updates are performed on this table
4. Updates are then copied to the source table
5. This procedure is alled twice, so before getting called for the second time, table is truncated.
Thanks,
AJ