Select with no conditions taking long time, too many blocks, pctused?

select * from at_journal;is taking 15 seconds, which is absurd. I have similar situations with other log tables in this system.
(selecting with rownum < 5 and with first_rows hint still takes 15 seconds)
Suspicious parameters:
PCT_FREE 1
PCT_USED 99
(Some other log tables have PCT_FREE 5, PCTUSED 90)
The tables have frequent inserts, once every day all records older than 90 days get deleted.
After computing statistics:
Blocks: 42,304 (seems absurdly high)
Size: 663 Mb
NUM_ROWS: 505,966
After copying data without compression to new table:
Blocks: 3,785
Size: 60 Mb
NUM_ROWS: 505,966
The select on the new table is instantaneous.
On a side note, in general is compression for a table with frequent deletes OK? It seems OK to me.
Storage parameters of table creation script:
PCTUSED    99
PCTFREE    1
INITRANS   1
MAXTRANS   255
STORAGE    (
            INITIAL          64K
            MINEXTENTS       1
            MAXEXTENTS       UNLIMITED
            PCTINCREASE      0
            FREELISTS        1
            FREELIST GROUPS  1
            BUFFER_POOL      DEFAULT
LOGGING
NOCOMPRESS
NOCACHE
NOPARALLEL
NOMONITORING;I imagine that most blocks for the table are empty. In my mind pctused of 99 means most of the block space should be full. I have read several explanations of PCTUSED/pctfree including oracle documentation and it seems pctused of 99 means the block should be 99% full.
Much thanks,
goo

goo wrote:
I imagine that most blocks for the table are empty. In my mind pctused of 99 means most of the block space should be full. I have read several explanations of PCTUSED/pctfree including oracle documentation and it seems pctused of 99 means the block should be 99% full.Hello, you don't give your database version, but if you were using Automatic Segment Space Management you would not need to worry about the PCTUSED setting, and it could be that you already are:
  1  select tablespace_name, segment_space_management
  2 from dba_TABLESPACES
TABLESPACE_NAME                SEGMEN
SYSTEM                         MANUAL
SYSAUX                         AUTO
UNDOTBS1                       MANUAL
TEMP                           MANUAL
USERS                          AUTO
EXAMPLE                        AUTOAnd if you are already using ASSM, then the PCTUSED is not an issue (PCTFREE still applies, but it's not an issue in your case).
goo wrote:
On a side note, in general is compression for a table with frequent deletes OK? It seems OK to me.Frequent DELETEs could be an issue because they will not reset the High Watermark, though there will not be a performance penalty because of the compression for DELETEs (the record only needs to be located for a DELETE, not decompressed). Your PCTUSED, however, is high enough so that any newly deleted record slots should be quickly reused, but... it depends on how those new records are being INSERTed. If, as Daniel says, the HWM is being effectively bypassed (Direct Path INSERTs, etc.), then the table will grow in size every day, but the volume of records effectively stays just about the same.
Have a look at how the new records are being added to that table, if it's a:
INSERT /*+ APPEND */ INTO tableA
SELECT...Then the HWM is being advanced, and the newly vacated slots from the DELETEd records are not being reused. So the table will continually grow, for roughly the same amount of data, and there'll be a proportionately longer data retrieval time.

Similar Messages

  • Nested table collection in select query "in clause" taking long time

    create or replace type t_circuitids is table of varchar2(100);
    --Below anonymous block keeps on running and never ends
    DECLARE
       v_circuitid    t_circuitids;
       v_count number;
       l_circuitids   VARCHAR2 (4000)
          := 'value1,value2,value3,value4,value5';
    BEGIN
    --Below query converts comma concatinated output to list and stores in nested table collection v_circuitids
       WITH a AS
            (SELECT ',' || l_circuitids || ',' AS circuitid
               FROM DUAL)
       SELECT DISTINCT TRIM (SUBSTR (circuitid,
                                     INSTR (circuitid, ',', 1, LEVEL) + 1,
                                       INSTR (circuitid, ',', 1, LEVEL + 1)
                                     - INSTR (circuitid, ',', 1, LEVEL)
                                     - 1
                            ) cid
       BULK COLLECT INTO v_circuitid
                  FROM a
            CONNECT BY LEVEL <
                             LENGTH (circuitid)
                             - LENGTH (REPLACE (circuitid, ','));
       SELECT COUNT (1)
         INTO v_count
         FROM table
        WHERE name IN (SELECT COLUMN_VALUE
                                          FROM TABLE (v_circuitid));
    END;
    --I got the issue, query "SELECT COLUMN_VALUE FROM TABLE (v_circuitid)" which is used in above code is responsible for this.
    --Same code works fine in Development and Test environments, But in prod it keeps on running
    --I solved this issue by creating a temp table, loading all values in collection into the temp table and using that temp table in "in clause"    "
    --Can any one answer why its behaving like this when i use collection in where clause?
    --I am using Oracle 9i

    Below is the code i used
    DECLARE
       v_circuitid    t_circuitids;
       v_count number;
       l_circuitids   VARCHAR2 (4000)
          := 'value1,value2,value3,value4,value5';
    BEGIN
    --Below query converts comma concatinated output to list and stores in nested table collection v_circuitids
       WITH a AS
            (SELECT ',' || l_circuitids || ',' AS circuitid
               FROM DUAL)
       SELECT DISTINCT TRIM (SUBSTR (circuitid,
                                     INSTR (circuitid, ',', 1, LEVEL) + 1,
                                       INSTR (circuitid, ',', 1, LEVEL + 1)
                                     - INSTR (circuitid, ',', 1, LEVEL)
                                     - 1
                            ) cid
       BULK COLLECT INTO v_circuitid
                  FROM a
            CONNECT BY LEVEL <
                             LENGTH (circuitid)
                             - LENGTH (REPLACE (circuitid, ','));
       SELECT COUNT (1)
         INTO v_count
         FROM table
        WHERE name IN (SELECT COLUMN_VALUE
                                          FROM TABLE (ccard(v_circuitid)));
    END;
    And got below error
    ORA-06550: line 27, column 5:
    PL/SQL: ORA-00906: missing left parenthesis
    ORA-06550: line 24, column 4:
    PL/SQL: SQL Statement ignored

  • Create table taking long time

    my select query(2m records) coming within a second but while creating a table (nologging) based on the select clause it is taking long time.
    can anybody give me the suggestion which part i will look to improve the performance..

    Plan
    SELECT STATEMENT ALL_ROWS Cost: 11 Bytes: 655 Cardinality: 1                                              
         19 FILTER                                         
              18 NESTED LOOPS Cost: 11 Bytes: 655 Cardinality: 1                                    
                   15 NESTED LOOPS Cost: 9 Bytes: 617 Cardinality: 1                               
                        12 NESTED LOOPS Cost: 8 Bytes: 481 Cardinality: 1                          
                             9 NESTED LOOPS Cost: 6 Bytes: 435 Cardinality: 1                     
                                  6 NESTED LOOPS Cost: 4 Bytes: 209 Cardinality: 1                
                                       3 TABLE ACCESS BY INDEX ROWID TABLE OYSTER_WEB3.TRANSACTION Cost: 2 Bytes: 155 Cardinality: 1           
                                            2 BITMAP CONVERSION TO ROWIDS      
                                                 1 BITMAP INDEX SINGLE VALUE INDEX (BITMAP) OYSTER_WEB3.IX_LINE_COMMODITY_ID
                                       5 TABLE ACCESS BY INDEX ROWID TABLE OYSTERPLUS_DATA.BRIO_SUPPLIERS Cost: 2 Bytes: 54 Cardinality: 1           
                                            4 INDEX UNIQUE SCAN INDEX (UNIQUE) OYSTERPLUS_DATA.PK_BRIO_SUPPLIERS Cost: 1 Cardinality: 1      
                                  8 TABLE ACCESS BY INDEX ROWID TABLE OYSTER3.FLAT_SITE_MV Cost: 2 Bytes: 226 Cardinality: 1                
                                       7 INDEX UNIQUE SCAN INDEX (UNIQUE) OYSTER3.PK_FLAT_SITE_MV Cost: 1 Cardinality: 1           
                             11 TABLE ACCESS BY INDEX ROWID TABLE OYSTER3.SITE_COMMODITY_CODING Cost: 2 Bytes: 46 Cardinality: 1                     
                                  10 INDEX UNIQUE SCAN INDEX (UNIQUE) OYSTER3.PK_SITE_COMMODITY_CODING Cost: 1 Cardinality: 1                
                        14 TABLE ACCESS BY INDEX ROWID TABLE OYSTERPLUS_DATA.BRIO_COMMODITIES Cost: 1 Bytes: 136 Cardinality: 1                          
                             13 INDEX UNIQUE SCAN INDEX (UNIQUE) OYSTERPLUS_DATA.PK_BRIO_COMMODITIES Cost: 0 Cardinality: 1                     
                   17 TABLE ACCESS BY INDEX ROWID TABLE OYSTER3.SUPPLIER_ALIAS Cost: 2 Bytes: 38 Cardinality: 1                               
                        16 INDEX UNIQUE SCAN INDEX (UNIQUE) OYSTER3.PK_SUPPLIER_ALIAS Cost: 1 Cardinality: 1

  • Hi in my sql query i applied like condition (like '%TEST') but it is taking long time

    Hi in my sql query i applied like condition (like '%TEST') but it is taking long time. I applied indexes also,but still i'm facing the same problem. In my databse nearly 2,00,000 records their.

    Hi Manikandan
    Is there a difference in performance between running the query in BEx and WebI?
    have you aggregates in place on the BEx side of things?
    When you say its taking too long to open the report, have you a variable screen coming up for the report and is that what is taking the time or is it the report execution.
    With regards
    Gill

  • Taking Long Time To Show Printer Selection Window

    Hi experts,
    When users are taking printout through SAP,it's taking long time to show printer selection window(Local Printer Selection Windows)
    present i am using sap ehp7 on sybase in windows.
    Thanks in advance..

    Hi,
    Have you tried by reinstalling of printer with patch updation of printer as well as of windows also ? If possible try to uninstall the printer & check again by reselection.
    Addition to it could you please share SAP GUI version on the system & confirm for, are you getting this issue on every system ( OS ) ?
    Regards,
    Gaurav

  • HT201250 my time capsule is taking too much time indexing backup and then taking longer time to back up ( 207 days ) or longer !!! what shall i do ?

    my time capsule is taking too much time indexing backup and then taking longer time to back up ( 207 days ) or longer !!! what shall i do ?

    Try 10.7.5 supplemental update.
    This update seems to have solved this problem for many.
    Best.

  • Problem with JSP taking long time to load.

    Hi,
    we have three server of oracle apps 10.1.3.
    only on one server we are getting problem.
    the problem is some jsp pages are taking long time to load and ssome times they dont evn load completly. half of the page is loaded and dont give any error message in opmn log.
    the same pages are working fyn in other servers .
    plesae help me.....

    This only happens in Photoshop 32 bit. Works fine in 64 bit.

  • Nested joins taking long time

    Hi Experts,
    Select query is taking long time execution.
    *- Get the Goods receipts  mainly selected per period (=> MKPF secondary
      SELECT mseg~ebeln mseg~ebelp mseg~werks
             ekko~bukrs ekko~lifnr ekko~zterm ekko~ekorg ekko~ekgrp
             ekko~inco1 ekko~exnum
             lfa1~name1 lfa1~land1 lfa1~ktokk lfa1~stceg
             mkpf~mblnr mkpf~mjahr mseg~zeile mkpf~bldat mkpf~budat
             mseg~bwart
    *Start of changes for CIP 6203752 by PGOX02
             mseg~smbln
    *End of changes for CIP 6203752 by PGOX02
             ekpo~matnr ekpo~txz01 ekpo~menge ekpo~meins
             ekbe~menge ekbe~dmbtr ekbe~wrbtr ekbe~waers
             ekpo~lgort ekpo~matkl ekpo~webaz ekpo~konnr ekpo~ktpnr
             ekpo~plifz ekpo~bstae
             INTO corresponding fields of TABLE it_temp
    *--Begin of modification
    *    FROM mkpf JOIN mseg ON mseg~mblnr EQ mkpf~mblnr
         FROM mkpf INNER JOIN mseg ON mseg~mandt EQ mkpf~mandt
                           and  mseg~mblnr EQ mkpf~mblnr
    *--End of modification
                           AND mseg~mjahr EQ mkpf~mjahr
                  JOIN ekbe ON ekbe~ebeln EQ mseg~ebeln
                           AND ekbe~ebelp EQ mseg~ebelp
                           AND ekbe~zekkn EQ '00'
                           AND ekbe~vgabe EQ '1'
                           AND ekbe~gjahr EQ mseg~mjahr
                           AND ekbe~belnr EQ mseg~mblnr
                           AND ekbe~buzei EQ mseg~zeile
                  JOIN ekpo ON ekpo~ebeln EQ ekbe~ebeln
                           AND ekpo~ebelp EQ ekbe~ebelp
                  JOIN ekko ON ekko~ebeln EQ ekpo~ebeln
                  JOIN lfa1 ON lfa1~lifnr EQ ekko~lifnr
              WHERE mkpf~budat IN so_budat
          AND mkpf~bldat IN so_bldat
          AND mkpf~vgart EQ 'WE'
          AND mseg~bwart IN so_bwart
          AND mseg~matnr IN so_matnr
          AND mseg~werks IN so_werks
          AND mseg~lifnr IN so_lifnr
          AND mseg~ebeln IN so_ebeln
          AND ekko~ekgrp IN so_ekgrp
          AND ekko~bukrs IN so_bukrs
          AND ekpo~matkl IN so_matkl
          AND ekko~bstyp IN so_bstyp
          AND ekpo~loekz EQ space
          AND ekpo~plifz IN so_plifz.
    in st05 it is showing MKPF is taking more time. Need ur sugesstions need ur help
    Moderator message - Please use code tags to format your code
    Edited by: Rob Burbank on Feb 5, 2010 9:21 AM

    Hi,
    your result set is quite big.
    20296 records in 283,5 seconds (13,9 ms per record) might not be optimal
    but definatelly the big result set will take time to execute, since you are reading data from 6 tables:
    SELECT
    FROM mkpf
    JOIN mseg ON u2026
    JOIN ekbe ON u2026
    JOIN ekpo ON u2026
    JOIN ekko ON u2026
    JOIN lfa1 ON u2026
    what run time do you expect?
    In case we would be able to get one record in 5ms it will still run more than 100 seconds.
    Generally: Limit the result set as far as possible (with where conditions), select only the columns that are really needed.
    Geneal nested loop join optimization:
    check if the optimizer chooses the table that has the smallest result set (take all where conditions for each table
    and use SE16 to count the result sets, the smallest one should be the start for the nested loop join.
    For the starting table the selective where condition fields (those the limit the result set significantly for that table) should be indexed.
    For other tables the join condition (and maybe additional selecitve where conditions for this table) should be indexed. but most likely you will join to the tables using the primary key which is indexed... .
    Kind regards,
    Hermann

  • SSRS Reports taking long time to load

    Hello,
    Problem : SSRS Reports taking long time to load
    My System environment : Visual Studio 2008 SP1  and SQL Server 2008 R2
    Production Environment : Visual Studio 2008 SP1  and SQL Server 2008 R2
    I have created a Parameterized report (6 parameters), it will fetch data from 1 table. table has 1 year and 6 months data,      I am selecting parameters for only 1 month (about 2500 records). It is taking almost 2 minutes and 30 seconds
    to load the report.
    This report running efficiently in my system (report load takes only 5 to 6 seconds) but in
    production it is taking 2 minutes 30 seconds.
    I have checked the Execution log from production so I found the timing for
    Data retrieval (approx~)       Processing (approx~)               Rendering (approx~)
    10 second                                      15 sec                        
                2 mins and 5 sec.
    But Confusing point is that , if I run the same report at different time overall output time is same (approx) 2 min 30 sec but
    Data retrieval (approx~)       Processing (approx~)                Rendering (approx~)
    more than 1 min                            15 sec                                     
    more than 1 min
    so 1 question why timings are different ?
    My doubts are
    1) If query(procedure to retrieve the data) is the problem then it should take more time always,
    2) If Report structure is problem then rendering will also take same time (long time)
    for this (2nd point) I checked on blog that Rendering depends on environment structure e.g. Network bandwidth, RAM, CPU Usage , Number of users accessing same report at a time.
    So I did testing of report when no other user working on any report But failed (same result  output is 2 min 30 sec)
    From network team I got the result is that there is no issue or overload in CPU usage or RAM also No issue in Network bandwidth.
    Production Database Server and Report server are different (but in same network).
    I checked that database server the SQL Server is using almost Full RAM (23 GB out of 24 GB)
    I tried to allocate the memory to less amount up to 2GB (Trial solution I got from Blogs) but this on also failed.
    one hint I got from colleague that , change the allocated memory setting from static memory to dynamic to SQL Server
    (I guess above point is the same) I could not find that option Static and Dynamic memory setting.
    I did below steps
    Connected to SQL Server Instance
    Right click on Instance go to properties, Go to Memory Tab
    I found three options 1) Server Memory   2) Other memory   3) Section for "Configured values and Running values"
    Then I tried to reduce Maximum  Server memory up to 2 GB (As mentioned above)
    All trials failed, this issue I could not find the roots for this issue.
    Can anyone please help (it's bit urgent).

    Hi UdayKGR,
    According to your description, your report takes too long to load on your production environment. Right?
    In this scenario, since the report runs quickly in developing environment, we initially think it supposed to be the issue on data retrieval. However, based on the information in execution log, it takes longest time on rendering part. So we suggest you optimize
    the report itself to reduce the time for rendering. Please refer to the link below:
    My report takes too long to render
    Here is another article about overall performance optimization for Reporting Services:
    Reporting Services Performance and Optimization
    If you have any question, please feel free to ask.
    Best Regards,
    Simon Hou

  • Connecting to the database taking long time to connect database server

    Hi
    When I execute procedure i am getting the below message at bottom of the Oracle SQL Developer
    "Connecting to the database"
    it is taking more than 10 min plz guide

    Hi
    have you installed a normal Oracle Client also on your Host? normal Oracle Client
    Did you connect with host:port:sid or with a Oracle Naming Service? through TNS Service
    Can you test tnsping <alias> yes, It is working fine
    Did other user have the same problem? yes
    Did you connect through WAN or LAN connection? LAN (Intranet)
    Can you tell more about you client/database setup?
    Database setup:
    OS: Window 2008 server
    version: 11.1.0
    Client: 11.1.0
    OS: Window 2008 server
    Now I am not able to execute single select query which table contains 6 records and 15 columns it is taking long time I have waited 30 min still no resutls
    only one table is behaving like this remaining is working fine
    Edited by: user9235224 on Oct 6, 2012 7:06 PM

  • The ODS activation is taking long time

    Hi,
    We are on SAP NetWeaver BI 701 (Support Package 5).
    We create a Z ODS, it will contain a lot of data (180.000.000 month-end) and we want to generate specific reports about it.
    The activation is taking long time, I assume is because we checked the flag "SIDs Generation upon Activation". I am confused about this check. I really need it? is this check the only problem.
    Thanks for you help.
    Victoria

    Hi Victoria:
       If your Z DSO is used only for staging purposes (you don't have queries based on this DSO and you send the data to another DSO or to an InfoCube) then you don't need to check the "SIDs Generation Upon Activation" box.
    Even more, to achieve better performance during data loads in this scenario, you might consider using a Write Optimized DSO instead of a Standard DSO, but if you decide to take this alternative don't forget to select the "Do Not check Uniqueness of Data" box if you need to write several records with the same Semantic Key.
    Regards,
    Francisco Milán.

  • F4 Help is taking long time

    Hi All,
    We are working on BI 7.0. version
    In the varaible pop-up screen we have two info objects.
    1. Fiscal year Period
    2. JOA(Joint operating aggriment)
    If u press F4 for JOA, it is taking long time to execute and finally the application is getting closed.same situation is there in RSRT also.
    If i enter with out JOA the query is giving the output. Here i have to restrict the query by JOA.
    i have changed the JOA peroperties in query designer.
    Query execution for filter value selection = Values in master data table.......
    but still the situation is the same.......
    Could you please suggest any solution for this.....
    Thanks & Regards,
    PK

    Hi Kamal,
    You can set that at the query level in the query designer for each query.
    1. Select the corresponding characteristic in the query designer.
    2. Goto to the "Extended tab" in the properties
    3. Select the "Values in the Master data table" in the "Query execution in the filter value selection.
    Also see some recomendations:
    Note 748623 - Input help (F4) has a very long runtime - recommendations
    Hope this helps.
    CK

  • Update ztable is taking long time

    Hi All,
    i have run the 5 jobs with the same program at a time but when we check the db trace
    zs01 is taking long time as shown below.here zs01 is having small amount of data.
    in the below dbtrace for updating zs01 is taking 2,315,485 seconds .how to reduce this?
    HH:MM:SS.MS Duration     Program   ObjectName  Op.   Curs   Array   Rec     RC     Conn     
    2:36:15 AM     2,315,485     SAPLZS01  ZS01       FETCH  294     1     1     0     R/3     
    The code as shown below
    you can check the code in the program SAPLZS01 include LZS01F01.
    FORM UPDATE_ZS01.
    IF ZS02-STATUS = '3'.
        IF Z_ZS02_STATUS = '3'.            "previous status is ERROR
          EXIT.
        ELSE.
          SELECT SINGLE FOR UPDATE * FROM  ZS01
                 WHERE  PROC_NUM    = ZS02-PROC_NUM.
          CHECK SY-SUBRC = 0.
          ADD ZS02-MF_AMT TO ZS01-ERR_AMT.
          ADD 1           TO ZS01-ERR_INVOI.
          UPDATE ZS01.
        ENDIF.
      ENDIF.
    my question is when updating the ztable why it is taking such long time,
    how to reduce the time or how to make faster to update the ztable .
    Thanks in advance,
    regards
    Suni

    Try the code like this..
    data: wa_zs01 type zs01.
    FORM UPDATE_ZS01.
    IF ZS02-STATUS = '3'.
        IF Z_ZS02_STATUS = '3'.            "previous status is ERROR
          EXIT.
        ELSE.
          SELECT SINGLE FOR UPDATE * FROM  ZS01
                 WHERE  PROC_NUM    = ZS02-PROC_NUM.
    -- change
      CHECK SY-SUBRC = 0.
          ADD ZS02-MF_AMT TO wa_ZS01-ERR_AMT.
          ADD 1           TO wa_ZS01-ERR_INVOI.
          UPDATE ZS01 from wa_zs01.
        ENDIF.
      ENDIF.
    And i think this Select query for ZS01 is inside the ZS02 SELECT statement,
    This might also make slow process.
    If you want to make database access always use Workarea/Internal table to fetch the data
    and work with that.
    Accessing database like this or with Select.... endselect is an inefficient programming.

  • Simple query is taking long time

    Hi Experts,
    The below query is taking long time.
    [code]SELECT   FS.*
      FROM   ORL.FAX_STAGE FS
             INNER JOIN
                   ORL.FAX_SOURCE FSRC
                INNER JOIN
                   GLOBAL_BU_MAPPING GBM
                ON GBM.BU_ID = FSRC.BUID
             ON UPPER (FSRC.FAX_NUMBER) = UPPER (FS.DESTINATION)
    WHERE       FSRC.IS_DELETED = 'N'
             AND GBM.BU_ID IS NOT NULL
             AND UPPER (FS.FAX_STATUS) ='COMPLETED';[/code]
    this query is returning 1645457 records.
    [code]PLAN_TABLE_OUTPUT
    | Id  | Operation           | Name                   | Rows  | Bytes | Cost (%CPU)|
    |   0 | SELECT STATEMENT    |                        |   625K|   341M| 45113   (1)|
    |   1 |  HASH JOIN          |                        |   625K|   341M| 45113   (1)|
    |   2 |   NESTED LOOPS      |                        |   611 | 14664 |    22   (0)|
    |   3 |    TABLE ACCESS FULL| FAX_SOURCE             |  2290 | 48090 |    22   (0)|
    |   4 |    INDEX RANGE SCAN | GLOBAL_BU_MAPPING_BUID |     1 |     3 |     0   (0)|
    |   5 |   TABLE ACCESS FULL | FAX_STAGE              |  2324K|  1214M| 45076   (1)|
    PLAN_TABLE_OUTPUT
    Note
       - 'PLAN_TABLE' is old version
    15 rows selected.[/code]
    The distinct number of records in each table.
    [code]SELECT FAX_STATUS,count(*)
    FROM fax_STAGE
    GROUP BY FAX_STATUS;
    FAX_STATUS    COUNT(*)
    BROKEN          10
    Broken - New    9
    Completed    2324493
    New             20
    SELECT is_deleted,COUNT(*)
    FROM  FAX_SOURCE
    GROUP BY IS_DELETED;
    IS_DELETED COUNT(*)
    N         2290
    Y         78[/code]
    Total number of records in each table.
    [code]SELECT COUNT(*) FROM ORL.FAX_SOURCE FSRC-- 2368
    SELECT COUNT(*) FROM ORL.FAX_STAGE--2324532
    SELECT COUNT(*) FROM APPS_GLOBAL.GLOBAL_BU_MAPPING--9
    [/code]
    To improve the performance of this query I have created the following indexes.
    [code]Functional based index on UPPER (FSRC.FAX_NUMBER) ,UPPER (FS.DESTINATION) and UPPER (FS.FAX_STATUS).
    Bitmap index on FSRC.IS_DELETED.
    Normal Index on GBM.BU_ID and FSRC.BUID.
    [/code]
    But still the performance is bad for this query.
    What can I do apart from this to improve the performance of this query.
    Please help me .
    Thanks in advance.

    <I have created the following indexes.
    CREATE INDEX ORL.IDX_DESTINATION_RAM ON ORL.FAX_STAGE(UPPER("DESTINATION"))
    CREATE INDEX ORL.IDX_FAX_STATUS_RAM ON ORL.FAX_STAGE(LOWER("FAX_STATUS"))
    CREATE INDEX ORL.IDX_UPPER_FAX_STATUS_RAM ON ORL.FAX_STAGE(UPPER("FAX_STATUS"))
    CREATE INDEX ORL.IDX_BUID_RAM ON ORL.FAX_SOURCE(BUID)
    CREATE INDEX ORL.IDX_FAX_NUMBER_RAM ON ORL.FAX_SOURCE(UPPER("FAX_NUMBER"))
    CREATE BITMAP INDEX ORL.IDX_IS_DELETED_RAM ON ORL.FAX_SOURCE(IS_DELETED)
    After creating the following indexes performance got improved.
    But our DBA said that new BITMAP index at FAX_SOURCE table (ORL.IDX_IS_DELETED_RAM) can cause locks
    on multiple rows if IS_DELETED column is in use. Please proceed with detailed tests.
    I am sending the explain plan before creating indexes and after indexes has been created.
    SELECT  FS.*
    FROM  ORL.FAX_STAGE FS
                    INNER JOIN
                    ORL.FAX_SOURCE FSRC
                  INNER JOIN
                      GLOBAL_BU_MAPPING GBM
                    ON GBM.BU_ID = FSRC.BUID
                ON UPPER (FSRC.FAX_NUMBER) = UPPER (FS.DESTINATION)
    WHERE      FSRC.IS_DELETED = 'N'
              AND GBM.BU_ID IS NOT NULL
              AND UPPER (FS.FAX_STATUS) =:B1;
    --OLD without indexes
    PLAN_TABLE_OUTPUT
    Plan hash value: 3076973749
    | Id  | Operation          | Name                  | Rows  | Bytes | Cost (%CPU)| Time    |
    |  0 | SELECT STATEMENT    |                        |  141K|    85M| 45130  (1)| 00:09:02 |
    |*  1 |  HASH JOIN          |                        |  141K|    85M| 45130  (1)| 00:09:02 |
    |  2 |  NESTED LOOPS      |                        |  611 | 18330 |    22  (0)| 00:00:01 |
    |*  3 |    TABLE ACCESS FULL| FAX_SOURCE            |  2290 | 59540 |    22  (0)| 00:00:01 |
    |*  4 |    INDEX RANGE SCAN | GLOBAL_BU_MAPPING_BUID |    1 |    4 |    0  (0)| 00:00:01 |
    |*  5 |  TABLE ACCESS FULL | FAX_STAGE              | 23245 |    13M| 45106  (1)| 00:09:02 |
    PLAN_TABLE_OUTPUT
    Predicate Information (identified by operation id):
      1 - access(UPPER("FSRC"."FAX_NUMBER")=UPPER("FS"."DESTINATION"))
      3 - filter("FSRC"."IS_DELETED"='N')
      4 - access("GBM"."BU_ID"="FSRC"."BUID")
          filter("GBM"."BU_ID" IS NOT NULL)
      5 - filter(UPPER("FS"."FAX_STATUS")=SYS_OP_C2C(:B1))
    21 rows selected.
    --NEW with indexes.
    PLAN_TABLE_OUTPUT
    Plan hash value: 665032407
    | Id  | Operation                        | Name                    | Rows  | Bytes | Cost (%CPU)| Time    |
    |  0 | SELECT STATEMENT                |                          |  5995 |  3986K|  3117  (1)| 00:00:38 |
    |*  1 |  HASH JOIN                      |                          |  5995 |  3986K|  3117  (1)| 00:00:38 |
    |  2 |  NESTED LOOPS                  |                          |  611 | 47658 |    20  (5)| 00:00:01 |
    |*  3 |    VIEW                          | index$_join$_002        |  2290 |  165K|    20  (5)| 00:00:01 |
    |*  4 |    HASH JOIN                    |                          |      |      |            |      |
    |*  5 |      HASH JOIN                  |                          |      |      |            |      |
    PLAN_TABLE_OUTPUT
    |  6 |      BITMAP CONVERSION TO ROWIDS|                          |  2290 |  165K|    1  (0)| 00:00:01 |
    |*  7 |        BITMAP INDEX SINGLE VALUE | IDX_IS_DELETED_RAM      |      |      |            |      |
    |  8 |      INDEX FAST FULL SCAN      | IDX_BUID_RAM            |  2290 |  165K|    8  (0)| 00:00:01 |
    |  9 |      INDEX FAST FULL SCAN        | IDX_FAX_NUMBER_RAM      |  2290 |  165K|    14  (0)| 00:00:01 |
    |* 10 |    INDEX RANGE SCAN              | GLOBAL_BU_MAPPING_BUID  |    1 |    4 |    0  (0)| 00:00:01 |
    |  11 |  TABLE ACCESS BY INDEX ROWID    | FAX_STAGE                | 23245 |    13M|  3096  (1)| 00:00:38 |
    |* 12 |    INDEX RANGE SCAN              | IDX_UPPER_FAX_STATUS_RAM |  9298 |      |  2434  (1)| 00:00:30 |
    Predicate Information (identified by operation id):
    PLAN_TABLE_OUTPUT
      1 - access(UPPER("DESTINATION")="FSRC"."SYS_NC00035$")
      3 - filter("FSRC"."IS_DELETED"='N')
      4 - access(ROWID=ROWID)
      5 - access(ROWID=ROWID)
      7 - access("FSRC"."IS_DELETED"='N')
      10 - access("GBM"."BU_ID"="FSRC"."BUID")
          filter("GBM"."BU_ID" IS NOT NULL)
      12 - access(UPPER("FAX_STATUS")=SYS_OP_C2C(:B1))
    31 rows selected
    Please confirm on the DBA comment.Is this bitmap index locks rows in my case.
    Thanks.>

  • Retry recovery action is taking long time to complete

    Hi,
    I am in the process of testing error handling for my BPEL process.
    As part of it, I have given the default recover action in the fault policy of the process as retry with a retry count of 3 times with retry interval as 2.
    To test this, i have changed the target end point uri of the target service to some invalid url and ran the process.
    First I was not able to find the BPEL process instance at all in the console.
    Later I was able to find the instance after half an hour. When I saw the instance, if retry had happened or not.
    I found that the retry happened for 3 times, but it took 30 mins for the first retry to complete and start the second retry.
    Can anyone please let me know why the retry is taking long time although the retry interval I gave is only 2 seconds.
    THanks in advance!!!!

    Hi,
    Please find below the fault policy file:
    <?xml version="1.0" encoding="UTF-8"?>
    <faultPolicy version="2.0.1" id="TestFaultPolicy" xmlns:env="http://schemas.xmlsoap.org/soap/envelope/" xmlns:xs="http://www.w3.org/2001/XMLSchema" xmlns="http://schemas.oracle.com/bpel/faultpolicy" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">
         <Conditions>
         <faultName>
         <condition>
         <action ref="aia-ora-retry"/>
         </condition>
         </faultName>
         </Conditions>
         <Actions>
              <Action id="aia-ora-retry">
                   <retry>
                        <retryCount>3</retryCount>
                        <retryInterval>2</retryInterval>
                        <exponentialBackoff/>
                        <retryFailureAction ref="aia-ora-java"/>
                        <retrySuccessAction ref="aia-ora-java"/>
                   </retry>
              </Action>
              <!-- This is an action will cause a replay scope fault-->
              <Action id="ora-replay-scope">
                   <replayScope/>
              </Action>
              <!-- This is an action will bubble up the fault-->
              <Action id="ora-rethrow-fault">
                   <rethrowFault/>
              </Action>
              <!-- This is an action will mark the work item to be "pending recovery from console"-->
              <Action id="ora-human-intervention">
                   <humanIntervention/>
              </Action>
              <!-- This action will cause the instance to terminate-->
              <Action id="ora-terminate">
                   <abort/>
              </Action>
              <Action id="aia-ora-java">
              <javaAction className="oracle.apps.aia.core.eh.BPELJavaAction" defaultAction="ora-rethrow-fault">
              <returnValue value="REPLAY" ref="ora-terminate"/>
              <returnValue value="RETRHOW" ref="ora-rethrow-fault"/>
              <returnValue value="ABORT" ref="ora-terminate"/>
              <returnValue value="RETRY" ref="aia-ora-retry"/>
              <returnValue value="MANUAL" ref="ora-human-intervention"/>
              </javaAction>
    </Action>     
         </Actions>
    </faultPolicy>

Maybe you are looking for

  • Dynamic scrollbar position?

    Hi all, I have a scenario where the user needs a list of materials that is sorted by material text, which is used as a web based sales order creation application. This list can be very long (1000+ lines) and is shown in a Table where one of the colum

  • Mail Sending- with Unicode enabled text in the attachment

    hi, i have a specific requirement for JAPAN country. its as follows: i am retrieving delivery details into my program & i need to send them as in a mail attachment(as CSV file) to particular mail id. for sending mail, i am using the function module "

  • WBS for two company codes

    Dear Experts, This requirement is for an IT service project. The process is - I have two company codes Comp code 1 & Comp code 2, In which Project is created in company code 1. Employees are assigned to project to onsite & offshore from company code

  • Zen Vision M static sounding while playi

    I have a 30gb Zen Vision M and I love the heck out of this thing! I've always been devoted to Creative like a fanboy. I've had a few problems with my Micros(plural, at least 2, both broke cause I'm clumsy) and my original MuVo. Anyways, I've dropped

  • BAsic settings needed in system crm-2007 in newely installed

    HI all   I have small quesry here.i Have a system newley installed crm-2007 can you all please tell me what settings do i need to do to start working with..either marketing or service. as per my knowledge what i know is...navigation bar profiles,busi