Fast refresh view Matt

Hi all

http://docs.Oracle.com/CD/B19306_01/server.102/b14223/basicmv.htm#i1006519

Fast refresh for a materialized view containing joins and aggregates is possible after any type of DML for the base tables (direct load or conventional INSERT , UPDATE , or DELETE ). It can be set to refresh ON COMMIT or ON DEMAND . A REFRESH ON COMMIT materialized view will be updated automatically when a transaction that is DML to one of the details of the materialized view commits tables. The time required to perform the validation may be slightly longer than usual when this method is selected. This is because the refresh operation is performed as part of the validation process. Therefore, this method may not be suitable if many users at the same time change the tables on which is based the materialized view.

1. I got this quote from the above link.

What is the refresh option by default when I use REFRESH QUICKLY? My example:

CREATE MATERIALIZED VIEW REFRESH QUICKLY ENABLE QUERY REWRITE AS mat_vw_test

SELECT *.

OF vw_test;

2 using the example 8-1 1 example: creating a materialized from the link view.

Product_sales_mv CREATE MATERIALIZED VIEW

REFRESH QUICKLY

ACTIVATE THE QUERY REWRITING

Did YOU SELECT p.prod_name, SUM (s.amount_sold) AS dollar_sales,

Count (*) AS cnt, COUNT (s.amount_sold) AS cnt_amt

S SALES, products p

WHERE s.prod_id = p.prod_id GROUP BY p.prod_name;

View of the mat will not be updated after a user execute INSERT, UPDATE, or UPDATE. fix?

How can I update the matte view after a DML statement is executed?

With the help of Oracle 11 g Release 11.2.0.4.0

Thanks in advance.

It was missing an index for a column in one of my big table.

Thanks for your help.

Best,

Tags: Database

Similar Questions

  • Failed to create the fast refresh on the materialized view

    Hi, I am creating a materialized on the tables with fast refresh view on commit but I get an error message.

    I have created the following logs:

    Bonus CREATE MATERIALIZED VIEW LOG ON
    Users TABLESPACE
    WITH THE PRIMARY KEY
    INCLUDING THE NEW VALUES

    Used to CREATE MATERIALIZED VIEW LOG ON
    Users TABLESPACE
    WITH THE PRIMARY KEY
    INCLUDING THE NEW VALUES

    I then try to create the following:

    Test.BAM CREATE MATERIALIZED VIEW

    IMMEDIATE CONSTRUCTION
    COOL OFF QUICKLY ON COMMIT
    AS
    SELECT a.dept_no, SUM (a.salary)
    EMPLOYEE a, b bonus
    WHERE a.employee_id = b.employee_id
    GROUP BY dept_no

    When I do that, I get the following error:

    . ORA-12032: cannot use column rowid of materialized

    view log on 'test '. "" BONUS ".

    If I change to a full refresh, it works pretty well. If I create a MV with a fast refresh on a query just within the limits of the table bonus it also works very well, for example ' select * bonus where employee_id = 2'

    What I am doing wrong? Also when looking at the notes for MV connects in some places, he said using the primary key and others, he says to the sequence of use and row_id. Which is correct and what difference does it make?

    11.2 aix.

    Thank you very much

    I suggest that investigate you the reason why your MV is not quickly updatable with the following method:

    (1) create MV_CAPABILITIES_TABLE with @?/rdbms/admin/utlxmv.sql

    (2) exec ('BAM') dbms_mview.explain_mview

    3)

    column format a100 msgtxt

    column related_text format a20

    SELECT the possible, capability_name, related_text,

    msgtxt

    OF mv_capabilities_table;

    That should tell you the reason why he isn't yet quickly updatable.

    There was a popular quote on men and fishing, but I don't remember it for now :-)

    Kind regards

    Uwe Hesse

  • The Materialized View - ORA-12052: is unable to fast refresh materialized view

    Hello

    I was hitting my head all day trying to create materialized views. I have made some progress, but have hit a brick wall, unfortunately!

    Basically, I have been invited to take a view and see if I can get the benefits of performance by transforming all or part of it in materialized views. Because the underlying tables of the config is updated quite frequently, I want to fast refresh materialized views on commit. However, when I try to create a materialized view containing an outer join in an aggregated materialized view table, I get ORA-12052: is unable to fast refresh materialized view.

    I went through the documentation and also Rob van Wijkvery useful series of blogs on the topic (especially http://rwijk.blogspot.co.uk/2009/09/fast-refreshable-materialized-view.html) but have not found anything that matches. Maybe I missed something somewhere along the line, or maybe I'm asking just something completely impossible?

    My db is 11.2.0.2.

    Here's the test scripts, I've worked with that:

    drop materialized view test1_test2_mv;
    Drop materialized view test2 journal;
    drop table test2;
    Drop materialized view test1_mv newspaper;
    drop materialized view test1_mv;
    Drop materialized view test1 journal;
    drop table test1;

    create table test1 (identification number,
    type varchar2 (10),
    number of Val,
    update_time date,
    constraint t1_pk primary key (id, type, val));

    Insert into test1
    Select 1, 'a', 1001, sysdate - 10/24 Union double all the
    Select 1, 'b', 1003, sysdate - 9/24 Union double all the
    Select 1, 'c', 1002, sysdate - 8/24 Union double all the
    Select 1, had ', 1004, sysdate - 7/24 Union double all the
    Select 1, 'e', 1005, sysdate - 6/24 Union double all the
    Select 1, 'c', 1006, sysdate - 5/24 Union double all the
    Select 2, 'a', 1002, sysdate - 4/24 Union double all the
    Select 2, 'b', 1005, sysdate - 3/24 Union double all the
    Select 3, 'a', 1001, sysdate - 2/24 Union double all the
    Select 3, 'c', 1006, sysdate - 1/24 Union double all the
    Select 3, 'e', 1008, sysdate - 2/24 Union double all the
    Select option 4, has ', 1004, sysdate - 3/24 Union double all the
    Select 5, 'b', 1002, sysdate - 4/24 Union double all the
    Select 5, 'g', 1001, sysdate - 5/24 Union double all the
    Select 6, 'h', 1004, sysdate - 6/24 Union double all the
    Select 7, 'b', 1007, sysdate - 7/24 Union double all the
    Select 7, had ', 1001, sysdate - 8/24 double;

    commit;

    Select * from test1;


    CREATE MATERIALIZED VIEW LOG ON test1
    WITH rowid, primary key (update_time)
    including the new values;


    Test1_mv CREATE MATERIALIZED VIEW
    IMMEDIATE CONSTRUCTION
    COOL OFF QUICKLY ON COMMIT
    Did YOU SELECT id,
    MAX (case when type = "there" end of val) THAT col_a,.
    MAX (case when type = 'b', then val end) AS col_b,.
    MAX (case when type = 'c' then end val) AS col_c,.
    MAX (case when type = ' then end of val) AS col_d,
    MAX (update_time) AS update_time
    OF test1
    WHERE TYPE in ('a',
    « b »,
    « c »,
    a ')
    GROUP BY id;

    CREATE MATERIALIZED VIEW LOG ON test1_mv
    WITH rowid
    including the new values;


    create table test2 (identification number,
    col2 number,
    COL3 varchar2 (10),
    number of COL4,
    constraint t2_pk primary key (id));

    Insert into test2
    Select 1, 1, 'bob', 1 double Union all
    Select 2, 1, "sue", 1 double Union all
    Select 3, 1, 'tom', 1 double Union all
    Select 4, 1, 'jay', 1 double Union all
    Select 5, 1, 'art', 1 double Union all
    Select 6, 1, 'kay', 1 double Union all
    Select 7, 1, 'max', 1 double Union all
    Select 8, 1, 'tim', 1 double Union all
    Select 9, 1, "liz", 1 from dual;

    commit;


    CREATE MATERIALIZED VIEW LOG ON test2
    WITH rowid, primary key
    including the new values;


    Test1_test2_mv CREATE MATERIALIZED VIEW
    IMMEDIATE CONSTRUCTION
    COOL OFF QUICKLY ON COMMIT
    LIKE SOME t2.rowid,.
    T1.ID,
    T1.col_a,
    T1.col_b,
    T1.col_c,
    T1.col_d,
    T1.update_time,
    T2.col2,
    T2. COL3
    OF test1_mv t1,.
    Test2 t2
    WHERE (+) t1.id = t2.id; -symbol of outer join is not correctly displayed on the forums without space, grr!

    ORA-12052: is unable to fast refresh materialized view TEST1_TEST2_MV

    Y at - it any way I can get the materialized view fast refresh on commit or I asking the impossible?

    Add t1.rowid:

    SQL > CREATE MATERIALIZED VIEW test1_test2_mv

    2 BUILD IMMEDIATE

    3 QUICK REFRESH YOU COMMIT

    4 AS t2.rowid SELECT rid2,

    5 t1.rowid rid1,

    6 t1.id

    T1.col_a 7,.

    T1.col_b 8,.

    T1.col_c 9,.

    T1.col_d 10,

    T1.update_time 11,

    T2.col2 12,

    13 t2.col3

    14 OF test1_mv t1,

    15 test2 T2

    16 WHERE t1.id = t2.id

    17.

    Materialized view created.

    SQL > select * from test1_test2_mv

    2.

    RID2 RID1 ID COL_A, COL_B, COL_C COL_D UPDATE_TIME COL2 COL3

    ------------------ ------------------ ---------- ---------- ---------- ---------- ---------- ------------------- ---------- ----------

    AAAYB6AANAAAANDAAA AAAYB/AANAAAAN/AAA 1 1001 1003 1006 1004 25 / 06 / 2014 12:54:16 1 bob

    AAAYB6AANAAAANDAAB AAAYB/AANAAAAN/AAB 2 1002 1005 25 / 06 / 2014 1 sue 14:54:16

    AAAYB6AANAAAANDAAC AAAYB/AANAAAAN/AAC 3 1001 1006 25 / 06 / 2014 16:54:16 1 tom

    AAAYB/AANAAAAN/AAD AAAYB6AANAAAANDAAD 4 1004 25/06/2014 14:54:16 1 jay

    AAAYB6AANAAAANDAAE AAAYB / AANAAAAN / AAE 5 1002 2014/06/25 13:54:16 1 art

    AAAYB6AANAAAANDAAF AAAYB/AANAAAAN/AAG 7 1007 1009 25 / 06 / 2014 10:54:16 max 1

    AAAYB/AANAAAAN/AAH                                                                                                        1 tim

    AAAYB/AANAAAAN/AAF                                                                                                        1 kay

    AAAYB/AANAAAAN/AAI                                                                                                        1 liz

    9 selected lines.

    SQL > update of test2

    2 set col3 = "fly."

    3 where id = 7

    6 m

    1 line update.

    SQL > validation

    2.

    Validation complete.

    SQL > select * from test1_test2_mv

    2.

    RID2 RID1 ID COL_A, COL_B, COL_C COL_D UPDATE_TIME COL2 COL3

    ------------------ ------------------ ---------- ---------- ---------- ---------- ---------- ------------------- ---------- ----------

    AAAYB6AANAAAANDAAA AAAYB/AANAAAAN/AAA 1 1001 1003 1006 1004 25 / 06 / 2014 12:54:16 1 bob

    AAAYB6AANAAAANDAAB AAAYB/AANAAAAN/AAB 2 1002 1005 25 / 06 / 2014 1 sue 14:54:16

    AAAYB6AANAAAANDAAC AAAYB/AANAAAAN/AAC 3 1001 1006 25 / 06 / 2014 16:54:16 1 tom

    AAAYB/AANAAAAN/AAD AAAYB6AANAAAANDAAD 4 1004 25/06/2014 14:54:16 1 jay

    AAAYB6AANAAAANDAAE AAAYB / AANAAAAN / AAE 5 1002 2014/06/25 13:54:16 1 art

    AAAYB/AANAAAAN/AAH                                                                                                        1 tim

    AAAYB/AANAAAAN/AAF                                                                                                        1 kay

    AAAYB/AANAAAAN/AAI                                                                                                        1 liz

    AAAYB6AANAAAANDAAF AAAYB/AANAAAAN/AAG 7 1007 1009 25 / 06 / 2014 10:54:16 1 rob

    9 selected lines.

  • Fast refresh of a materialized view works incorrect

    Hi all

    I created a MV using the sql below:

    create a materialized view WLS

    Refresh quickly

    activate the query rewriting

    as

    SELECT Employees.USER_NAME,

    Permissions.ROLE_NAME,

    Vouchers.VOUCHER_ID,

    EmployeeRoles.FROM_DATE,

    EmployeeRoles.TO_DATE,

    VoucherPatterns.rowid ACC_VOUCHER_PATTERNS_ROW_ID,

    Vouchers.ROWID ACC_VOUCHERS_ROW_ID,

    Ledgers.ROWID ACC_LEDGERS_ROW_ID,

    Employees.ROWID BANK_EMPLOYEES_ROW_ID,

    IncExcUnits.rowid INC_EXC_UNITS_ROW_ID,

    EmployeeRoles.rowid BANK_EMPLOYEE_ROLES_ROW_ID,

    ParentRoles.rowid PARENT_ROLES_ROW_ID,

    Roles.ROWID ROLES_ROW_ID,

    Permissions.ROWID PERMISSIONS_ROLES_ROW_ID,

    RolesMemberships.rowid ROLES_MEMBERSHIPS_ROW_ID

    OF ACC_VOUCHER_PATTERNS VoucherPatterns,.

    Good ACC_VOUCHERS,

    ACC_LEDGERS books,

    BANK_EMPLOYEES employees,

    BANK_EMPLOYEE_INC_EXC_UNITS IncExcUnits,

    BANK_EMPLOYEE_ROLES EmployeeRoles,

    ParentRoles ROLES,

    Roles,

    Permissions of ROLES,

    ROLES_MEMBERSHIPS RolesMemberships

    WHERE Vouchers.TYPE_VOUCHER_PATTERN_FK =

    VoucherPatterns.VOUCHER_PATTERN_ID

    AND VoucherPatterns.CONFIRM_COUNT = 2

    AND Vouchers.STATE = 'UIS.

    AND Vouchers.LEDGER_FK = Ledgers.LEDGER_ID

    AND Employees.ID = IncExcUnits.BKEM_ID (+)

    AND (((IncExcUnits.INC_EXCLUD_TYPE EST NULL ou))

    IncExcUnits.INC_EXCLUD_TYPE! = "EXC") AND

    Ledgers.ORGANIZATION_UNIT_FK = Employees.ORGU_ID) OR

    (Ledgers.ORGANIZATION_UNIT_FK = IncExcUnits.ORGU_ID AND)

    IncExcUnits.INC_EXCLUD_TYPE = 'INC'))

    AND EmployeeRoles.BKEM_ID = Employees.ID

    AND ParentRoles.ID = EmployeeRoles.ROLS_ID

    AND ParentRoles.IS_PERMISSION = n

    AND ParentRoles.ROLE_TYPE = 'ORG '.

    AND ParentRoles.ID = Roles.ROLS_ID

    AND Roles.IS_PERMISSION = n

    AND Roles.ROLE_TYPE = 'SYS '.

    AND Permissions.IS_PERMISSION = 'Y '.

    AND Permissions.ROLE_NAME = 'ACC_AccVoucherConfirm. '

    AND Roles.ID = RolesMemberships.ROLS_ID

    AND Permissions.ID = RolesMemberships.ROLS_ID_MEMBER_OF;

    Then I issued a SELECT statement * although it and that works very well.

    And then I manipulated the data in tables such as, according to the clause WHERE MV, it should return no rows.

    Intuitively it will return the last data placed on it, before updating the MV

    But it returns the same data after issuing the following command in PL/SQL Developer:

    Call dbms_mview.refresh ("list = > ' WLS", method = > 'F');

    And unfortunately what follows will fix:

    Call dbms_mview.refresh ("list = > ' WLS", method = > "C");

    Why the Quick cooling of a refreshable quick MV works fine when its full refresh works well?

    How ever after that defeat made changes to tables and deliver the fast refresh using the following command the MV will contain accurate updated data:

    Call dbms_mview.refresh ("list = > ' WLS", method = > 'F');

    It seems that there is bug in clear lines of Fast refreshable MV but upserts work very well.

    I tried to simulate the problem to a simple query with simple tables with sample data, but those who have no problem.

    Please help me,

    Thank you

    And the solution is simple,

    Just avoid using the OR condition described in the WHERE clause and use UNION ALL instead!

    Don't forget that you have a calculated column of unique Creator for each of your unions.

    Thanks to myself!

  • Materialized view conducive to fast refresh method

    In AWM when I select refresh FASTER method and select the materialized view, get the error below:
    "Quick Refresh method requires the materialized view logs and a complete refresh previously run the MV cube." I need this fast a cube materialized view refresh, so that it performs an incremental refresh and re-aggregation of only the changed lines in the source table.
    can someone help me on this?

    Remember to choose the 'Force' that runs fast if possible update, as otherwise it performs a complete refresh. You can also specify Complete as the Refresh method in the dialog box Assistant Maintenance first, you keep the cube.

    You'll need create materialized on all the base tables, view logs if you have not already done so, otherwise the cube will be able to do a fast refresh (differential).

  • Pivot can be used in fast refresh materialized view?

    Hi, I have a question about materialized view nested mode fast refresh on Oracle 11 g (It support function of pivot, but oracle 10 g doesn't support).

    When I created, he throws "ORA-12015: cannot create a view fast refresh materialized by a complex query.
    Then I used dbms_mview.explain_mview to see reason, and it tell me the following
    REFRESH_FAST_AFTER_INSERT ' view inline or subquery in LIST not supported for this type of MV.

    Can someone help me, any suggestions will be appreciated
    create table empX as select * from scott.emp;
alter table empX add constraint PK_empX_empno primary key (empno);

--drop  MATERIALIZED VIEW LOG ON empX;
CREATE MATERIALIZED VIEW LOG ON empX with rowid, sequence(empno);

--drop MATERIALIZED VIEW mv_empX;
CREATE MATERIALIZED VIEW mv_empX
REFRESH FAST START WITH SYSDATE
NEXT  SYSDATE + 1/1440
AS   
  select * from
  (
   select rowid emp_rowid, deptno, job, sal from empX
  )
  PIVOT( max(sal) for job IN ('ANALYST' job1, 'CLERK' job2, 'MANAGER' job3));
 
--select * from mv_capabilities_table
declare
  lv_sqltext varchar2(4000);
begin
  execute immediate 'truncate table mv_capabilities_table';
  lv_sqltext := 'select * from
  (
   select deptno, job, sal from empX
  )
  PIVOT( max(sal) for job IN (''ANALYST'' job1, ''CLERK'' job2, ''MANAGER'' job3))
  ';  
  dbms_mview.explain_mview(lv_sqltext,'nested=>TRUE');
  commit;
end;
/

    Let me help you...
    I have done following and it run/view updated because I just replaced quickly with a complete word, because there are limitations for quickly updatable views:

    SQL> show user;
USER is "SCOTT"
SQL> create table empX as select * from scott.emp;

Table created.

SQL> alter table empX add constraint PK_empX_empno primary key (empno);

Table altered.

SQL>
SQL> --drop  MATERIALIZED VIEW LOG ON empX;
SQL> CREATE MATERIALIZED VIEW LOG ON empX with rowid, sequence(empno);

Materialized view log created.

SQL>
SQL> --drop MATERIALIZED VIEW mv_empX;
SQL> CREATE MATERIALIZED VIEW mv_empX
  2  REFRESH COMPLETE WITH rowid
  3  START WITH sysdate
  4  NEXT  SYSDATE + 1/1440
  5  AS
  6    select * from
  7    (
  8     select rowid emp_rowid, deptno, job, sal from empX
  9    )
 10    PIVOT( max(sal) for job IN ('ANALYST' job1, 'CLERK' job2, 'MANAGER' job3));

Materialized view created.

SQL> select * from mv_empx;

EMP_ROWID              DEPTNO       JOB1       JOB2       JOB3
------------------ ---------- ---------- ---------- ----------
AAAShcAAEAAAATTAAN         10                  1300
AAAShcAAEAAAATTAAE         30
AAAShcAAEAAAATTAAJ         30
AAAShcAAEAAAATTAAC         30
AAAShcAAEAAAATTAAA         20                   800
AAAShcAAEAAAATTAAK         20                  1100
AAAShcAAEAAAATTAAM         20       3000
AAAShcAAEAAAATTAAD         20                             2975
AAAShcAAEAAAATTAAB         30
AAAShcAAEAAAATTAAI         10
AAAShcAAEAAAATTAAL         30                   950

EMP_ROWID              DEPTNO       JOB1       JOB2       JOB3
------------------ ---------- ---------- ---------- ----------
AAAShcAAEAAAATTAAF         30                             2850
AAAShcAAEAAAATTAAG         10                             2450
AAAShcAAEAAAATTAAH         20       3000

14 rows selected.

SQL> begin
  2  dbms_mview.refresh('SCOTT.mv_empx');
  3  end;
  4  /

PL/SQL procedure successfully completed.

SQL> select * from mv_empx;

EMP_ROWID              DEPTNO       JOB1       JOB2       JOB3
------------------ ---------- ---------- ---------- ----------
AAAShcAAEAAAATTAAN         10                  1300
AAAShcAAEAAAATTAAE         30
AAAShcAAEAAAATTAAJ         30
AAAShcAAEAAAATTAAC         30
AAAShcAAEAAAATTAAA         20                   800
AAAShcAAEAAAATTAAK         20                  1100
AAAShcAAEAAAATTAAM         20       3000
AAAShcAAEAAAATTAAD         20                             2975
AAAShcAAEAAAATTAAB         30
AAAShcAAEAAAATTAAI         10
AAAShcAAEAAAATTAAL         30                   950

EMP_ROWID              DEPTNO       JOB1       JOB2       JOB3
------------------ ---------- ---------- ---------- ----------
AAAShcAAEAAAATTAAF         30                             2850
AAAShcAAEAAAATTAAG         10                             2450
AAAShcAAEAAAATTAAH         20       3000

14 rows selected.

SQL>

    So, the answer is Yes, we can use Pivot with Materialized view but:
    1 MV must be full refresh.
    2 oracle version should be 11 g; of course the pivot is available in Oracle 11 g.

    If this answers your question, please close the message, otherwise continue.

    These links may also be of interest:
    http://docs.Oracle.com/CD/B28359_01/server.111/b28313/basicmv.htm#i1007028
    http://docs.Oracle.com/CD/B28359_01/server.111/b28313/basicmv.htm#i1007007

    http://rwijk.blogspot.in/2009/06/fast-refreshable-MATERIALIZED-view.html

    And:

    Please see if (Note: 179466.1 - view to fast refresh materialized diagnose ORA-12015 / complex queries) help.

    Concerning
    Girish Sharma

  • How to fast refresh materialized view parallel

    We have Oracle 11.2.0.1 on Redhat 5.2. We use a dozen views fast refresh materialized in our application.
    Occasionally refresh MV takes longer than the refresh interval.

    One of the ways to improve the performance of refresh MV is to update at the same time, for example
    http://www.dba-oracle.com/t_materialized_views_refreshing_performance.htm

- Use parallel DML - Oracle author Michael Armstrong Smith notes "I've done parallel materialized view refreshing
 on tables recently and improved the load times considerably. Rather than having one load which took 2 hours,
 I run 4 parallel loads, one for each partition. The length of time for the whole process is now determined by
 how long the biggest partition takes to load. In my case, this is 40 minutes, with two 30 minute loads and one
 20 minute load.  Overall I am saving 1 hour 20 minutes. I can now add further partitions and do the same thing.
 My only limitation is the parallel loads because I don't have unlimited processing power.
{code} My master tables are not partitioned, but think parallel will still better than non-parallel.

Now the question, how to make refresh parallel. According to the paragraph cited above, Use parallel DML appears doing the job. Can someone  confirm that.

We use {code}
alter materialied view MV_OFFENSE parallel REFRESH FAST start with sysdate next sysdate+ 1/24;
{code}to schedule refresh. Does the parallel clause here parallels query the MV or refresh the MV?

Any other way to make MV refresh parallel?

Thanks                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                               

    Hello

    Just search for * 577870.1*. But before you search, select ID of item in the source list.

    Hope this helps.
    Best regards
    Jean Valentine

  • Fast refresh materialized views

    Hi all

    We have tables in the remote database and we created materialized views in our local database using the database but link to refresh these take a long time, so we can use fast refresh on these materialized views?

    Kind regards
    RAM

    Look at ATOMIC_REFRESH option, if you set this to FALSE, you can see the performance because it will use the path of DIRECT ACCESS and TRUNCATE. Data will be unavailable while being refreshed well.

    See you soon
    If

  • Materialized - view fast refresh connects

    I understand that there is a data change log maintained internally to fast refresh materialized view (CDC) Capture occurs.

    My question is, will this newspaper get purged once the changes are applied to a corresponding materialized view or it will persist with other changes is attached to it?

    What would be the size of the change log for 150 million documents covering 40 base tables?

    Sorry, I haven't used groups of refreshment MV.

  • Fast refresh materialized view problems

    I have two databases (A and B).

    In A database, user NICK has a table called COLOUR_MASTER.
    Also in this database, NICK issues creating materialized command newspaper view colour_master with the primary key, including the new values (and Yes, there is a primary key defined for this table)
    In database B, there is a link of database called A_LINK, connect, Nick, identified by password using 'dbA ';
    In database B, user IAN also creates a materialized view CM_MV cool off quickly in select * from colour_master@A_LINK

    This statement is done correctly, and there are several million lines in CM_MV when it's done.

    Immediately, IAN issued this command:

    Start
    dbms_mview. Refresh ('CM_MV');
    end;

    .. .and after a small pause, it gets the error

    ORA-12008: error path refresh materialized view
    ORA-02068: following sever error of A_LINK
    ORA-03113: end of file on communication channel

    One more + "select count (*) of colour_master@A_LINK" + subject immediately manages to return the correct number of records in the database so A 3113 is a bit misleading, I think that, in this case have crash, the database remains accessible at all times, there is no network problem (database A and B are both located in the same server so it of not like I'm trying of) connect to halfway around the globe).

    The only thing I can think is that there is a permissions issue causes this error. The creation of MV because is made via a link, it is not a problem of IAN see all appropriate, but once the updates come into play, it is perhaps.

    Any who can shed light on why I can't do the fast refresh from the view, I can create joyfully, please?

    I'm surprised no one asked you to do what is obvious: try to do a select * from colour_master@A_LINK, without that being wrapped in a create materialized MATERIALIZED view. Just do a simple select * from...

    My bet is that you will get an error ORA_22992 can not use selected from the remote tables LOB Locators . And if you can't do a simple select * in a remote table, I suspect that you can reasonably expect Fireworks when you try to create or refresh a view materialized on such a thing!

    Here's a quick test:

    In the 1 database:
    =========

    SQL> create table A (col1 varchar2(2), col2 clob);
Table created.

SQL> alter table A add constraint A1 primary key (col1);
Table altered.

SQL> insert into A values ('AA','This is an entry into a clob column');
1 row created.

SQL> commit;
Commit complete.

SQL> create materialized view log on A with primary key including new values;
Materialized view log created.

    In the database 2
    =========

    SQL> connect ims_global/v0yager1@ussd
Connected.

SQL> create materialized view B refresh fast as
  2  select * from A@remotedb_link;

Materialized view created.

SQL> exec dbms_mview.refresh('B');
BEGIN dbms_mview.refresh('B'); END;

*
ERROR at line 1:
ORA-12008: error in materialized view refresh path
ORA-02068: following severe error from REMOTEDB_LINK
ORA-03113: end-of-file on communication channel
ORA-06512: at "SYS.DBMS_SNAPSHOT", line 2251
ORA-06512: at "SYS.DBMS_SNAPSHOT", line 2457
ORA-06512: at "SYS.DBMS_SNAPSHOT", line 2426
ORA-06512: at line 1

SQL> drop materialized view B;
Materialized view dropped.

SQL> create materialized view B refresh fast as
  2  select col1 from A@ims_link;

Materialized view created.

SQL> exec dbms_mview.refresh('B');
PL/SQL procedure successfully completed.

    Which seems to reproduce your problem rather well, I think.

    I don't know if there is no work around for this. I mean, obviously you can't miss the CLOB column in your statement of 'create a MV' as I just did, but I don't know if you're still allowed or expected to be able to create a MV on lob columns and that such and such correction will allow you to do. Maybe someone else can provide this information.

    But at least you won't have to play Russian roulette with your init.ora parameters!

  • View Matt

    Hi all

    I have a materialized view of a view of two tables tableA and tableB. I want to add a new column of another table many, I get this error:

    ORA-12052: is unable to fast refresh materialized view

    DROP MATERIALIZED VIEW LOG ON tableA;
DROP MATERIALIZED VIEW LOG ON tableB;
DROP MATERIALIZED VIEW LOG ON tableN;

CREATE MATERIALIZED VIEW LOG ON tableA WITH ROWID (tableA_id, fieldA), SEQUENCE INCLUDING NEW VALUES;
CREATE MATERIALIZED VIEW LOG ON tableB WITH ROWID (tableB_id, fieldB), SEQUENCE INCLUDING NEW VALUES;
CREATE MATERIALIZED VIEW LOG ON tableN WITH ROWID (tableN_id, fieldN), SEQUENCE INCLUDING NEW VALUES;


CREATE OR REPLACE VIEW vw_tablesAB AS
(SELECT a.fieldA, b.fieldB, n.fieldN
from tableA a, tableB b, newTable n
where a.idA = b.idA
and a.myId = n.myID);


CREATE MATERIALIZED VIEW mat_vw_tablesAB REFRESH FAST ON COMMIT WITH ROWID AS
SELECT *
FROM vw_tablesAB;

    The Manyfield of many have NULL values, that is the question?

    Did I miss something for quick update?

    Thanks in advance.

    Sorry, it took me a while to remember what I did before. I managed to do work as declaring the ROWID on the view as:

    1. VIEW to CREATE or REPLACE vw_tablesAB AS
    2. (SELECT
    3. a.ROWID as arowid,
    4. b.ROWID as browid,
    5. n.ROWID as nrowis,
    6. a.fieldA, b.fieldB, n.fieldN
    7. from tableA a, tableB b, NewTable n
    8. where a.idA = b.idA
    9. and a.myId = n.myID);

    Creating the matte view works fine. Thanks for your help!

    Best.

  • a unique index or unique constraint on the issue of view Matt

    10.2.0.3

    I have an OLTP table and a matte view to fast refresh of the table in the warehouse. I have unique indexes on the matte view just as I have on the OLTP table. Of course, it's a bad idea because the updating Oracle on mattress views mechanism does not apply to the dml in the same order that it occurred on the side of OLTP? Should I get rid of all the unique indexes on views mattress in my warehouse and create regular index because of their unique nature will just happen because the side OLTP has a unique index? What will be the impact on the performance of the queries? Here's the alert log...

    Journal of owp2 alerts
    =======================
    ORA-12012: error on auto work 1595
    ORA-12008: error path refresh materialized view
    ORA-00001: unique constraint (SMS_AR. IU02_ROUTE_REF_MRKR) violated
    ORA-06512: at "SYS." DBMS_SNAPSHOT", line 510
    ORA-06512: at line 1
    ORA-00001: unique constraint (SMS_AR. IU02_ROUTE_REF_MRKR) violated
    ORA-00610: internal error Code
    ORA-12012: error on auto work 260282
    ORA-30439: updating of the ' ORA-30439: updating of the 'SMS_AR MV_ROUTE_REF_MRKR' failed due to the ORA-12008: error in the path of refresh materialized view
    ORA-00001: unique constraint (SMS_AR. IU02_ROUTE_REF_MRKR) violated
    ORA-06512: at "SYS." DBMS_SNAPSHOT", line 2254
    ORA-06512: at "SYS." DBMS_SNAPSHOT", line 2460
    ORA-06512: at "SYS." DBMS_SNAPSHOT", line 2429
    ORA-06512: at "SMS_AR.PA_PIES_WAREHOUSE", line 44
    ORA-06512: at line 2
    ORA-20000: index 'SMS_AR '. "' I01_MV_PIES_INV_REFMKR ' or the partition of this index is unusable

    Mark Reichman wrote:
    I think that this problem is resolved... Unless someone has something else to add. I have not tried yet... But it seems to be valid. I did a test and a unique constraint can be delayed in fact creates a non-unique index. So I need to remove my unique index on my matte view and create unique constraints can be delayed.

    Or...

    forget the unqiue part and simply change the indexes not unique because the main table has a unique index and guarantees uniqueness for me and the matte view will simply copy whats in the main table.

    Mark,

    the solution seems reasonable. Just a note: If you use a "reportable" unique constraint Oracle ignores any attempt to perform a direct-path insert of access and still stations conventional insert generating undo and many more again.

    As long as you do only a 'rapid' refresh, it should not matter, but in case you deal with refreshs full large MVs, this could make a difference when running not atomic refreshs (who can take advantage of the direct-path inserts / DML etc. at the same time)...

    Then you can consider using only non-unique index rather than the constraint may be delayed if performance can matter and given the fact that you should never see duplicates in the MV because of the constraint on the base table.

    Furthermore, you can use a non-unique index to apply a not reportable unique/primary key constraint as well, it is supported. You just need to create the index yourself before you set the constraint or using explicit syntax "CREATE INDEX" of the constraint clause.

    For more information, I wrote a note on this problem "may be deferred" some time ago:

    http://Oracle-Randolf.blogspot.com/2008/07/Deferrable-constraints-and-direct-path.html

    Kind regards
    Randolf

    Oracle related blog stuff:
    http://Oracle-Randolf.blogspot.com/

    SQLTools ++ for Oracle (Open source Oracle GUI for Windows):
    http://www.sqltools-plusplus.org:7676 /.
    http://sourceforge.NET/projects/SQLT-pp/

  • materialized wiev fast refresh every minute

    Hello, I need create mv and this mv need auto fast refresh every minute.

    I first create the table 3

    first:

    CREATE TABLE 'HR '. "' EMPLOYEES '.

    (NUMBER (6.0) "EMPLOYE_ID",)

    VARCHAR2 (20 BYTE) "FIRST NAME",

    VARCHAR2 (25 BYTE) "NAME."

    VARCHAR2 (25 BYTE) "E-MAIL."

    "PHONE_NUMBER" VARCHAR2 (20 BYTE),

    'HIRE_DATE' DATE,

    "JOB_ID' VARCHAR2 (10 BYTE),

    "SALARY" NUMBER (8,2).

    NUMBER (2.2) "COMMISSION_PCT."

    NUMBER (6.0) "MANAGER_ID."

    "DEPARTMENT_ID" NUMBER (4.0)

    )

    Second:

    CREATE TABLE 'HR '. "" DEPARTMENTS. "

    ("DEPARTMENT_ID" NUMBER (4,0),)

    VARCHAR2 (30 BYTE) "DEPARTMENT_NAME."

    NUMBER (6.0) "MANAGER_ID."

    NUMBER (4,0) 'LOCATION_ID '.

    )

    Thirdly:

    CREATE TABLE 'HR '. "' LOCATIONS '.

    (NUMBER (4,0) 'LOCATION_ID',)

    'ADRESSE_RUE' VARCHAR2 (40 BYTE),

    "ZIP_CODE" VARCHAR2 (12-BYTE),

    VARCHAR2 (30 BYTE) 'CITY ',.

    VARCHAR2 (25 BYTE) "STATE_PROVINCE"

    'COUNTRY_ID' TANK (2 BYTES)

    )

    can I create mv newspaper with rowid, become I refresh mv with quick setting

    create log view materialized on EMPLOYEES with rowid.

    create log view materialized on departments with rowid.

    create log view materialized on the premises with rowid.

    Finally, I create mv with fast refresh method

    CREATE THE TEST_MV_1 MATERIALIZED VIEW

    REFRESH QUICKLY START WITH sysdate + 0 NEXT (sysdate + 1/1440)

    AS

    Select EMPLOYEES. EMPLOYEE_id

    departments.department_id

    locations.location_id

    EMPLOYEES.rowid EMPLOYEES

    departments.rowid departments

    locations.rowid locations

    EMPLOYEES

    ministries

    places

    where departments.manager_id (+) = EMPLOYEES. EMPLOYEE_ID

    and departments.location_id = locations.location_id (+);

    now I have my mv and I'm seeing or my mv running

    SELECT

    table_name,

    TO_CHAR (LAST_REFRESH,'yyyy-MM-DD HH24:MI:SS) LAST_REFRESH

    , to_char (sysdate,'yyyy-MM-DD HH24:MI:SS) sys_date

    -Count (Name)

    Of all_snapshots

    where table_name = 'TEST_MV_1 '.

    I get:

    TABLE-NAME LAST_REFRESH SYS_DATE

    ------------------------------ ------------------- -------------------

    TEST_MV_1 2015-09-01 17:56:10-2015-09-01 17:57:02

    After 2 min

    SELECT

    table_name,

    TO_CHAR (LAST_REFRESH,'yyyy-MM-DD HH24:MI:SS) LAST_REFRESH

    , to_char (sysdate,'yyyy-MM-DD HH24:MI:SS) sys_date

    -Count (Name)

    Of all_snapshots

    where table_name = 'TEST_MV_1 '.

    TABLE-NAME LAST_REFRESH SYS_DATE

    ------------------------------ ------------------- -------------------

    TEST_MV_1 2015-09-01 17:58:10 2015-09-01 17:59:05

    then I insert the new record in the table

    Insert employees

    (EMPLOYE_ID,

    FIRST NAME,

    LAST_NAME,

    E-mail

    PHONE_NUMBER,

    HIRE_DATE,

    JOB_ID,

    SALARY,

    COMMISSION_PCT,

    MANAGER_ID,

    DEPARTMENT_ID)

    values

    (515, 'Jennifer', 'Whalen', 'JWHALEN', '515.123.4444', to_date('1987.09.17','yyyy.mm.dd'), 'AD_ASST', null, 101, 10, 24000)

    /

    Commit

    /

    1 row inserted.

    Validation complete.

    and check again my last_refresh mv

    SELECT

    table_name,

    TO_CHAR (LAST_REFRESH,'yyyy-MM-DD HH24:MI:SS) LAST_REFRESH

    , to_char (sysdate,'yyyy-MM-DD HH24:MI:SS) sys_date

    -Count (Name)

    Of all_snapshots

    where table_name = 'TEST_MV_1 '.

    TABLE-NAME LAST_REFRESH SYS_DATE

    ------------------------------ ------------------- -------------------

    TEST_MV_1 2015-09-01 17:59:10 -2015-09-01 17:59:30

    affter 2 min

    TABLE-NAME LAST_REFRESH SYS_DATE

    ------------------------------ ------------------- -------------------

    TEST_MV_1 2015-09-01 17:59:10 -2015-09-01 18:02:12

    and when I go to data mv I don't see my new data

    Select * from TEST_MV_1

    where EMPLOYEE_ID = 515

    no selected line

    Maybe someone can help me? I'm two weeks try fix it and lost all hope


    I find:

    My mv refreshed, but when I insert new data he collapsed

    I ' am am ansver: when I try to refresh the MV with fuction

    DBMS_REFRESH. Refresh (the "HO".) "TEST_MV_1" ');

    I get the error

    ORA-12008: error path refresh materialized view

    ORA-00942: table or view does not exist

    ORA-06512: at "SYS." DBMS_SNAPSHOT", line 2563

    ORA-06512: at "SYS." DBMS_SNAPSHOT", line 2776

    ORA-06512: at "SYS." DBMS_IREFRESH', line 685

    ORA-06512: at "SYS." DBMS_REFRESH", line 195

    ORA-06512: at line 2

    00000 - "error in the path of refresh materialized view.

    * Cause: Table SNAP$ _ reads lines of sight

    MVIEW$ _, which is a view of the main table

    (the master can be on a remote site).  Any

    error in this way will cause this error when updating.

    For updates of the Rapids, the table. MLOG$ _

    is also referenced.

    * Action: Examine the other messages on the stack to find the problem.

    See if SNAP objects $ _, MVIEW$ _,.

    . @, . MLOG$ _@.

    There is always.

    so I think that I can not select the HR data. MLOG$ _EMPLOYEES and another newspaper in the HR schema table

    When I try to select * HR. MLOG$ _EMPLOYEES I get table emty (schema HO)

    so I don't understand how I can not update quick MV when I log empty.

    I therefore ask my db admin for more privileges to select the MV journal table

    I do not know what admin Preobrajensky give to my diet, but it's a job.

    dude thanks for the help

  • Listagg works does not in mview with fast refresh on commit?

    Version Info:

    Oracle Database 11 g Release 11.2.0.3.0 - 64 bit Production

    PL/SQL Release 11.2.0.3.0 - Production

    CORE Production 11.2.0.3.0

    AMT for 64-bit Windows: Version 11.2.0.3.0 - Production

    NLSRTL Version 11.2.0.3.0 - Production

    I have a materialized view that is defined as fast refresh on validation and includes a call to listagg. The update goes as planned, but the value of the column that is created by using the listagg not updated unless I have make an explicit refresh. I was able to recreate the problem with a simple case, pasted below. Someone knows how to fix this?

    Thank you!

    -Tom

    Test scenario follows...

    SQL> create table codes (
  2       id number primary key,
  3       product varchar2(10) not null,
  4       code varchar2(10) not null);

Table created.

SQL> create materialized view log on codes with rowid, sequence (product, code) including new values;

Materialized view log created.

SQL> create materialized view code_agg_mv
  2        refresh fast on commit
  3        with rowid
  4        as
  5  select c.product
  6        , listagg(c.code, ',') within group (order by c.code) codes
  7        , count(*) cnt
  8    from codes c
  9   group by c.product;

Materialized view created.

SQL> truncate table mv_capabilities_table;

Table truncated.

SQL> begin
  2       dbms_mview.explain_mview('CODE_AGG_MV');
  3       commit;
  4  end;
  5  /

PL/SQL procedure successfully completed.

SQL> select capability_name, possible
  2    from mv_capabilities_table
  3   where capability_name like '%REFRESH%'
  4   order by capability_name;

CAPABILITY_NAME                P
------------------------------ -
REFRESH_COMPLETE               Y
REFRESH_FAST                   Y
REFRESH_FAST_AFTER_ANY_DML     Y
REFRESH_FAST_AFTER_INSERT      Y
REFRESH_FAST_AFTER_ONETAB_DML  Y
REFRESH_FAST_PCT               N

6 rows selected.

SQL> insert into codes (id, product, code) values (1, 'Pants', 'Large');

1 row created.

SQL> insert into codes (id, product, code) values (2, 'Pants', 'Blue');

1 row created.

SQL> insert into codes (id, product, code) values (3, 'Pants', 'Itchy');

1 row created.

SQL> commit;

Commit complete.

SQL> select *
  2    from code_agg_mv;

PRODUCT    CODES                                 CNT
---------- ------------------------------ ----------
Pants      Blue,Itchy,Large                        3

SQL> -- ^^^ This is correct ^^^

SQL> delete from codes where id = 3;

1 row deleted.

SQL> commit;

Commit complete.

SQL> select * from code_agg_mv;

PRODUCT    CODES                                 CNT
---------- ------------------------------ ----------
Pants      Blue,Itchy,Large                        2

SQL> -- ^^^ CNT is correct, but codes should not contain "Itchy" ^^^
SQL> --
SQL> -- Try an explicit fast refresh...

SQL> begin
  2       dbms_mview.refresh('CODE_AGG_MV', 'F');
  3  end;
  4  /

PL/SQL procedure successfully completed.

SQL> select * from code_agg_mv;

PRODUCT    CODES                                 CNT
---------- ------------------------------ ----------
Pants      Blue,Itchy,Large                        2

SQL> -- ^^^ Nope, same issue. ^^^
SQL> --
SQL> -- Try an explicit complete refresh...

SQL> begin
  2       dbms_mview.refresh('CODE_AGG_MV', 'C');
  3  end;
  4  /

PL/SQL procedure successfully completed.

SQL> select * from code_agg_mv;

PRODUCT    CODES                                 CNT
---------- ------------------------------ ----------
Pants      Blue,Large                              2

SQL> -- ^^^ There we go. ^^^

    Message edited by Tom N to include database version information.

    Hello Tom,

    listagg() is not supported for fast refresh.

    http://docs.Oracle.com/CD/E11882_01/server.112/e25554/basicmv.htm#i1007028

    Only SUM, COUNT, AVG, STDDEV, VARIANCE, MIN and MAX are supported to fast refresh.

    Kind regards

    Pravin

  • Commit performance on table with Fast Refresh MV

    Hello world

    Try to wrap your head around fast refresh performance and why I see (what I consider) high disk numbers / query associated with the update of the MV_LOG in a TKPROF.

    The installation program.
    (Oracle 10.2.0.4.0)

    Database table:
    SQL> desc action;
 Name                                      Null?    Type
 ----------------------------------------- -------- ----------------------------
 PK_ACTION_ID                              NOT NULL NUMBER(10)
 CATEGORY                                           VARCHAR2(20)
 INT_DESCRIPTION                                    VARCHAR2(4000)
 EXT_DESCRIPTION                                    VARCHAR2(4000)
 ACTION_TITLE                              NOT NULL VARCHAR2(400)
 CALL_DURATION                                      VARCHAR2(6)
 DATE_OPENED                               NOT NULL DATE
 CONTRACT                                           VARCHAR2(100)
 SOFTWARE_SUMMARY                                   VARCHAR2(2000)
 MACHINE_NAME                                       VARCHAR2(25)
 BILLING_STATUS                                     VARCHAR2(15)
 ACTION_NUMBER                                      NUMBER(3)
 THIRD_PARTY_NAME                                   VARCHAR2(25)
 MAILED_TO                                          VARCHAR2(400)
 FK_CONTACT_ID                                      NUMBER(10)
 FK_EMPLOYEE_ID                            NOT NULL NUMBER(10)
 FK_ISSUE_ID                               NOT NULL NUMBER(10)
 STATUS                                             VARCHAR2(80)
 PRIORITY                                           NUMBER(1)
 EMAILED_CUSTOMER                                   TIMESTAMP(6) WITH LOCAL TIME
                                                     ZONE


SQL> select count(*) from action;

  COUNT(*)
----------
   1388780
    MV was created
    create materialized view log on action with sequence, rowid
(pk_action_id, fk_issue_id, date_opened) 
including new values;

-- Create materialized view
create materialized view issue_open_mv
build immediate
refresh fast on commit
enable query rewrite as 
select  fk_issue_id issue_id,
     count(*) cnt,
     min(date_opened) issue_open,
     max(date_opened) last_action_date,
     min(pk_action_id) first_action_id,
     max(pk_action_id) last_action_id,
     count(pk_action_id) num_actions
from    action
group by fk_issue_id;

exec dbms_stats.gather_table_stats('tg','issue_open_mv')

SQL> select table_name, last_analyzed from dba_tables where table_name = 'ISSUE_OPEN_MV';

TABLE_NAME                     LAST_ANAL
------------------------------ ---------
ISSUE_OPEN_MV                  15-NOV-10

*note: table was created a couple of days ago *

SQL> exec dbms_mview.explain_mview('TG.ISSUE_OPEN_MV');

CAPABILITY_NAME                P REL_TEXT MSGTXT
------------------------------ - -------- ------------------------------------------------------------
PCT                            N
REFRESH_COMPLETE               Y
REFRESH_FAST                   Y
REWRITE                        Y
PCT_TABLE                      N ACTION   relation is not a partitioned table
REFRESH_FAST_AFTER_INSERT      Y
REFRESH_FAST_AFTER_ANY_DML     Y
REFRESH_FAST_PCT               N          PCT is not possible on any of the detail tables in the mater
REWRITE_FULL_TEXT_MATCH        Y
REWRITE_PARTIAL_TEXT_MATCH     Y
REWRITE_GENERAL                Y
REWRITE_PCT                    N          general rewrite is not possible or PCT is not possible on an
PCT_TABLE_REWRITE              N ACTION   relation is not a partitioned table

13 rows selected.
    Fast refresh works fine. And the newspaper is kept small enough.
    SQL> select count(*) from mlog$_action;

  COUNT(*)
----------
         0
    When I update a row in the base table:
    var in_action_id number;

exec :in_action_id := 398385;

UPDATE action
SET emailed_customer = SYSTIMESTAMP
WHERE pk_action_id = :in_action_id
AND DECODE(emailed_customer, NULL, 0, 1) = 0
/

commit;
    What follows, I get via tkprof.
    ********************************************************************************

INSERT /*+ IDX(0) */ INTO "TG"."MLOG$_ACTION" (dmltype$$,old_new$$,snaptime$$,
  change_vector$$,sequence$$,m_row$$,"PK_ACTION_ID","DATE_OPENED",
  "FK_ISSUE_ID")
VALUES
 (:d,:o,to_date('4000-01-01:00:00:00','YYYY-MM-DD:HH24:MI:SS'),:c,
  sys.cdc_rsid_seq$.nextval,:m,:1,:2,:3)


call     count       cpu    elapsed       disk      query    current        rows
------- ------  -------- ---------- ---------- ---------- ----------  ----------
Parse        1      0.00       0.01          0          0          0           0
Execute      2      0.00       0.03          4          4          4           2
Fetch        0      0.00       0.00          0          0          0           0
------- ------  -------- ---------- ---------- ---------- ----------  ----------
total        3      0.00       0.04          4          4          4           2

Misses in library cache during parse: 1
Misses in library cache during execute: 1
Optimizer mode: CHOOSE
Parsing user id: SYS   (recursive depth: 1)

Rows     Row Source Operation
-------  ---------------------------------------------------
      2  SEQUENCE  CDC_RSID_SEQ$ (cr=0 pr=0 pw=0 time=28 us)


Elapsed times include waiting on following events:
  Event waited on                             Times   Max. Wait  Total Waited
  ----------------------------------------   Waited  ----------  ------------
  db file sequential read                         4        0.01          0.01
********************************************************************************

********************************************************************************

update "TG"."MLOG$_ACTION" set snaptime$$ = :1
where
 snaptime$$ > to_date('2100-01-01:00:00:00','YYYY-MM-DD:HH24:MI:SS')


call     count       cpu    elapsed       disk      query    current        rows
------- ------  -------- ---------- ---------- ---------- ----------  ----------
Parse        1      0.00       0.01          0          0          0           0
Execute      1      0.94       5.36      55996      56012          1           2
Fetch        0      0.00       0.00          0          0          0           0
------- ------  -------- ---------- ---------- ---------- ----------  ----------
total        2      0.94       5.38      55996      56012          1           2

Misses in library cache during parse: 1
Misses in library cache during execute: 1
Optimizer mode: CHOOSE
Parsing user id: SYS   (recursive depth: 1)

Rows     Row Source Operation
-------  ---------------------------------------------------
      0  UPDATE  MLOG$_ACTION (cr=56012 pr=55996 pw=0 time=5364554 us)
      2   TABLE ACCESS FULL MLOG$_ACTION (cr=56012 pr=55996 pw=0 time=46756 us)


Elapsed times include waiting on following events:
  Event waited on                             Times   Max. Wait  Total Waited
  ----------------------------------------   Waited  ----------  ------------
  db file scattered read                       3529        0.02          4.91
********************************************************************************

select dmltype$$, max(snaptime$$)
from
 "TG"."MLOG$_ACTION"  where snaptime$$ <= :1  group by dmltype$$


call     count       cpu    elapsed       disk      query    current        rows
------- ------  -------- ---------- ---------- ---------- ----------  ----------
Parse        1      0.00       0.00          0          0          0           0
Execute      1      0.00       0.00          0          0          0           0
Fetch        2      0.70       0.68      55996      56012          0           1
------- ------  -------- ---------- ---------- ---------- ----------  ----------
total        4      0.70       0.68      55996      56012          0           1

Misses in library cache during parse: 1
Misses in library cache during execute: 1
Optimizer mode: CHOOSE
Parsing user id: SYS   (recursive depth: 1)

Rows     Row Source Operation
-------  ---------------------------------------------------
      1  SORT GROUP BY (cr=56012 pr=55996 pw=0 time=685671 us)
      2   TABLE ACCESS FULL MLOG$_ACTION (cr=56012 pr=55996 pw=0 time=1851 us)


Elapsed times include waiting on following events:
  Event waited on                             Times   Max. Wait  Total Waited
  ----------------------------------------   Waited  ----------  ------------
  db file scattered read                       3529        0.00          0.38
********************************************************************************

delete from "TG"."MLOG$_ACTION"
where
 snaptime$$ <= :1


call     count       cpu    elapsed       disk      query    current        rows
------- ------  -------- ---------- ---------- ---------- ----------  ----------
Parse        1      0.00       0.00          0          0          0           0
Execute      1      0.71       0.70      55946      56012          3           2
Fetch        0      0.00       0.00          0          0          0           0
------- ------  -------- ---------- ---------- ---------- ----------  ----------
total        2      0.71       0.70      55946      56012          3           2

Misses in library cache during parse: 1
Misses in library cache during execute: 1
Optimizer mode: CHOOSE
Parsing user id: SYS   (recursive depth: 1)

Rows     Row Source Operation
-------  ---------------------------------------------------
      0  DELETE  MLOG$_ACTION (cr=56012 pr=55946 pw=0 time=702813 us)
      2   TABLE ACCESS FULL MLOG$_ACTION (cr=56012 pr=55946 pw=0 time=1814 us)


Elapsed times include waiting on following events:
  Event waited on                             Times   Max. Wait  Total Waited
  ----------------------------------------   Waited  ----------  ------------
  db file scattered read                       3530        0.00          0.39
  db file sequential read                        33        0.00          0.00
********************************************************************************
    Could someone explain why the the SELECT/UPDATE/DELETE on MLOG$ _ACTION if 'expensive' when it should be only 2 rows (the old value and the new value) in this newspaper after an update? I could do to improve the performance of the update?

    Let me know if you need more info... would be happy to provide.

    My guess would be that you were once a very large transaction that inserted a large number of rows in this table. So the table segment is big enough now and the high watermark is average at the end of this segment, causing a full scan table to analyze a large number of empty blocks and retrieve the two lines.

    You can issue a truncation on this table of $ MLOG: which would free up the empty blocks and brings back the high-watermark in the first block.

Maybe you are looking for