Scheduling of SOA purge instances (11.1.1.7) generating logs

Question

What is best practice for scheduling the purging of SOA instances using the Oracle Soa purge strategy?
Followed the instructions successfully http://docs.oracle.com/cd/E29542_01/admin.1111/e10226/soaadmin_partition.htm#SOAAG97268
My requirements:
For each purge run a unique log must be generated on the filesystem.
The purging must be done frequently according to a schedule (i.e. daily 8pm).
My issue:
when I run the looped purge (soa.delete_instances), logs are not being generated as expected within the SOA_PURGE_DIR.
logs are only generated in SOA_PURGE_DIR when I execute the parallel purge (soa.delete_instances_in_parallel). I need to run looped purge and not parallel purge.
Current scenario (workaround to generate logs (seems a bit overkill)):
DBMS_Scheduler job invoking a shell script
Shell script invokes a custom sql script spooling the output to a log file
Custom sql script invokes the Oracle soa.delete_instances procedure setting all the parameters
soa.delete_instances runs.
Preferred scenario (not working):DBMS_Scheduler job (PL/SQL block) -> Oracle SOA Purge procedure
DBMS_Scheduler:
BEGIN
dbms_scheduler.create_job('SOA_PURGE',
    job_type             =>'EXECUTABLE',
    job_action           =>'/stage/scripts/sh/soa_purge.sh',
    number_of_arguments  =>0,
    start_date           =>TO_TIMESTAMP_TZ('27-NOV-2014 08.00.00.000000000 PM +01:00','DD-MON-RRRR HH.MI.SSXFF AM TZR','NLS_DATE_LANGUAGE=english'),
    repeat_interval      =>'FREQ=DAILY',
    end_date             =>NULL,
    job_class            =>'DEFAULT_JOB_CLASS',
    enabled              =>TRUE,
    auto_drop            =>FALSE,
    comments             =>'Job to purge the data from dehydration gateway database.'
COMMIT;
END;
Shell script:
SPOOL_FILE1=/stage/logs/soa_purge_$(date +%Y_%M_%DT%H_%M_%S).log
sqlplus "user/password" <<EOF
alter session set nls_date_format = 'dd/mm/yyyy hh24:mi:ss';
SET linesize 100
SET pagesize 300
SET time on
SET timing on
spool ${SPOOL_FILE1}
@@/stage/scripts/sql/debug_on.sql
@@/stage/scripts/sql/soa_purge_test.sql
@@/stage/scripts/sql/debug_off.sql
spool off;
set time off;
exit;
EOF
Custom script (soa_purge_test.sql (above)):
SET SERVEROUTPUT ON;
DECLARE
  MAX_CREATION_DATE TIMESTAMP;
  MIN_CREATION_DATE TIMESTAMP;
  BATCH_SIZE        INTEGER;
  MAX_RUNTIME       INTEGER;
  RETENTION_PERIOD  TIMESTAMP;
BEGIN
  MIN_CREATION_DATE := TO_TIMESTAMP(TO_CHAR(sysdate-90, 'YYYY-MM-DD'),'YYYY-MM-DD');
  MAX_CREATION_DATE := TO_TIMESTAMP(TO_CHAR(sysdate-30, 'YYYY-MM-DD'),'YYYY-MM-DD');
  RETENTION_PERIOD  := TO_TIMESTAMP(TO_CHAR(sysdate-29, 'YYYY-MM-DD'),'YYYY-MM-DD');
  MAX_RUNTIME       := 1380;
  BATCH_SIZE        := 250000;
  SOA.DELETE_INSTANCES(
    MIN_CREATION_DATE    => MIN_CREATION_DATE,
    MAX_CREATION_DATE    => MAX_CREATION_DATE,
    BATCH_SIZE           => BATCH_SIZE,
    MAX_RUNTIME          => MAX_RUNTIME,
    RETENTION_PERIOD     => RETENTION_PERIOD
END;

Answer 1

What is best practice for scheduling the purging of SOA instances using the Oracle Soa purge strategy?
Followed the instructions successfully http://docs.oracle.com/cd/E29542_01/admin.1111/e10226/soaadmin_partition.htm#SOAAG97268
My requirements:
For each purge run a unique log must be generated on the filesystem.
The purging must be done frequently according to a schedule (i.e. daily 8pm).
My issue:
when I run the looped purge (soa.delete_instances), logs are not being generated as expected within the SOA_PURGE_DIR.
logs are only generated in SOA_PURGE_DIR when I execute the parallel purge (soa.delete_instances_in_parallel). I need to run looped purge and not parallel purge.
Current scenario (workaround to generate logs (seems a bit overkill)):
DBMS_Scheduler job invoking a shell script
Shell script invokes a custom sql script spooling the output to a log file
Custom sql script invokes the Oracle soa.delete_instances procedure setting all the parameters
soa.delete_instances runs.
Preferred scenario (not working):DBMS_Scheduler job (PL/SQL block) -> Oracle SOA Purge procedure
DBMS_Scheduler:
BEGIN
dbms_scheduler.create_job('SOA_PURGE',
    job_type             =>'EXECUTABLE',
    job_action           =>'/stage/scripts/sh/soa_purge.sh',
    number_of_arguments  =>0,
    start_date           =>TO_TIMESTAMP_TZ('27-NOV-2014 08.00.00.000000000 PM +01:00','DD-MON-RRRR HH.MI.SSXFF AM TZR','NLS_DATE_LANGUAGE=english'),
    repeat_interval      =>'FREQ=DAILY',
    end_date             =>NULL,
    job_class            =>'DEFAULT_JOB_CLASS',
    enabled              =>TRUE,
    auto_drop            =>FALSE,
    comments             =>'Job to purge the data from dehydration gateway database.'
COMMIT;
END;
Shell script:
SPOOL_FILE1=/stage/logs/soa_purge_$(date +%Y_%M_%DT%H_%M_%S).log
sqlplus "user/password" <<EOF
alter session set nls_date_format = 'dd/mm/yyyy hh24:mi:ss';
SET linesize 100
SET pagesize 300
SET time on
SET timing on
spool ${SPOOL_FILE1}
@@/stage/scripts/sql/debug_on.sql
@@/stage/scripts/sql/soa_purge_test.sql
@@/stage/scripts/sql/debug_off.sql
spool off;
set time off;
exit;
EOF
Custom script (soa_purge_test.sql (above)):
SET SERVEROUTPUT ON;
DECLARE
  MAX_CREATION_DATE TIMESTAMP;
  MIN_CREATION_DATE TIMESTAMP;
  BATCH_SIZE        INTEGER;
  MAX_RUNTIME       INTEGER;
  RETENTION_PERIOD  TIMESTAMP;
BEGIN
  MIN_CREATION_DATE := TO_TIMESTAMP(TO_CHAR(sysdate-90, 'YYYY-MM-DD'),'YYYY-MM-DD');
  MAX_CREATION_DATE := TO_TIMESTAMP(TO_CHAR(sysdate-30, 'YYYY-MM-DD'),'YYYY-MM-DD');
  RETENTION_PERIOD  := TO_TIMESTAMP(TO_CHAR(sysdate-29, 'YYYY-MM-DD'),'YYYY-MM-DD');
  MAX_RUNTIME       := 1380;
  BATCH_SIZE        := 250000;
  SOA.DELETE_INSTANCES(
    MIN_CREATION_DATE    => MIN_CREATION_DATE,
    MAX_CREATION_DATE    => MAX_CREATION_DATE,
    BATCH_SIZE           => BATCH_SIZE,
    MAX_RUNTIME          => MAX_RUNTIME,
    RETENTION_PERIOD     => RETENTION_PERIOD
END;