Similar Questions

• To reset a bad windows update

  I stopped a update Windows junk 10 months ago and still get this code. Updated for Windows 10 failure error WindowsUpdate_C1900101-20003 when I go into my control panel and look for news updates for my (beloved) Windows 7. How can I erase this upward? I thank in advance.

  [Fusion]

  How to erase a bad windows update? Is there a Temp file or download that I can remove? I love Windows 7 but I have a window 10 day that I stopped and you want that he erased!

  You seem to be on W7 and tries to clear a update that maybe wanted to take you to W10.  Your question does not belong to this site, because this site makes testing beta versions of future versions of W10.

  However, if you go on the d/l from Microsoft site and search for "wushowhide.diagcab", it is a program that allows you to hide updates.  I have used in the past.  Let it run for a few moments and it can offer hide this update unnecessary and probably for an offer that is no longer valid.

  Mel

• ORA-54017: excluded from the regime of virtual columns UPDATE Operation

  Referenced to

  ORA-54017: excluded from the regime of virtual columns UPDATE Operation

  There are problems with v4.0_4.121.1.0 knew?

  Thank you

  Sorry, but after setting StoreGeneratedPattern to Computed in the EF model I have the same problem.

  Thank you

• How to check the column updated whereby the package or procedure?

  Hi all

  Can someone help me how to check the column updated by which the package or procedure

  A.Mahesh

  Hello.

  You can check what object is a reference to the table of the column:

  Select *.

  of all_dependencies d

  where d.REFERENCED_NAME = '. '

  And you can find all the links to the object source data column:

  Select *.

  of s all_source

  where s.name = "."

  and s.OWNER = '. '

  and s.TYPE = "."

  and as s.TEXT '%%'

• Bad column index

  Hello
  
  When I start my replicat I got the following

  Bad column index (10) specified for table SIEBEL.S_RESP_BU, max columns = 10

  I have more than 2 columns in most target tables (other columns are the same).
  What is the pb?
  and if so what can I do?
  
  Thanks for your help
  
  My replicat setting is below:

  replicat rep1
  setenv (oracle_sid=XXX)
  userid ogg,password xxxx
  handlecollisions
  assumetargetdefs
  discardfile /tst/data/ogg/discards/resp.dsc,purge
  reperror default discard
  tableexclude siebel.S_SRM_TASK_HIST
  map SIEBEL.*, target SIEBEL.*;

  Check that your tables are identical. Currently, they are not, and by using ASSUMETARGETDEFS is not valid.

• Rendering fonts were going badly after update 32.0

  After the recent update, the police started making bad. Kind of blur with jagged edges and transparent parts. It used to be well D:

  Sometimes it is OK for a while I scroll pages, and sometimes when I stop scrolling, the police remains as it is. as if there are specific positions render text correctly. However, not all policies are affected. Those inside the scrollable area.

  Times that the police properly, return the bookmarks is, title bar and the tab name is unaffected and appears is not smooth. Same is not the case on other browsers. All my other system fonts look good; It's really just Firefox.

  Restart Firefox = failed
  Restart computer = failed
  Tried the clear type = setting did not work
  Tried to activate and disable hardware acceleration on Firefox = failed
  Tried the activation and the deactivation of the graphics card itself = failed
  Tried pretty much all the advice I found on Google = none worked
  Tried the complete uninstall and re - install Firefox 32.0 = did not work

  Graphics driver is updated.
  Windows is updated.

  I use Windows 7 Ultimate 32 bit
  Graphics AMD Radeon HD 6800 card

  I fixed it by downgrading to Firefox 31 who assures me that 32.0 is the culprit. As I was about to send the message at the top. I tried without decommissioning as my last resort, and it worked. Now I send you only that in a report. I hope you will understand what caused it because I'm quite strict (OC) that everything in my PC must stay update > _ <. so, I want to use Firefox 32.0 as much as possible.

  You can try with the catalyst beta driver of http://support.amd.com/en-us/kb-artic.../latest-catalyst-windows-beta.aspx? 14.4 catalyst has been blocklisted for direct2d because of stability issues in firefox 32 (bug #984488).

• IOS 9.3.1 bad news updated

  The new 9.3.1 update is bad and does not solve the problem with the links

  When you click on some links on facebook, the facebook app close and get out

  Please solve this problem

  Yo the need for people who can hear to help you. Click on the link below.

  http://www.Apple.com/feedback/

• Satellite A60-202 after bad BIOS updated the PC does not work

  Hello

  After a very bad my BIOS update, my PC doesn't work and doesn't turn on.
  What to do now?
  It is Toshiba Satellite A60-202 with 512 MB of memory and 40 GB of HD.
  model PSA60E-OLP06TFR SA60-202

  Thank you

  Yves

  HI Yves,

  Unfortunately, if the BIOS update procedure does not work (or fill), then it is likely that the BIOS in the CMOS chip had been corrupt and without it, you will not be able to boot your laptop to fix the problem.

  It will be necessary to bring (or send) your laptop to a repairer authorized to have reflashed BIOS.

  Kind regards

• How can I fix a bad HP update file?

  I'm having a problem with my laptop computer dv6701us under Vista. It freezes a few times and my AVG virus program will stop on a bad file, I switched to Avast! and I have the same problem. I ran Malwarebytes and it hangs up on the following file:

  "C:\PROGRAM Hewlett-Packard HP SOFTWARE UI\PC UPDATES\FRONT END\fi\ui\STYLES\HP\default.css"

  How doI identify the update and remove/replace this update?  For most machine works very well in the case otherwise, I ran 'disk health check' and the HD is fine.  I switched to Avast A / V AVG 9.0 because wants to make a "business optimization" and never to complete this analysis. Very frustrating to deal with.  Someone at - it a similar experience?  Help?

  Orudy1

  Go to the dashboard, programs & features and uninstall HP Update. Reboot and try your scan. If you have success, you can install the latest HP Software (if desired):

  FTP://ftp.HP.com/pub/SoftPaq/sp38501-39000/sp38874.exe

• Flag value does not change after the SCD TYPE1 column update

  Hello

  I had scd type3, type2 successfully.

  I'm trying to implement the type 1 scd.

  For this, I change slow dimension & ckm oracle the oracle select ikm.

  the source and target tables are the following. for the type 1 scd, I chose scd behavior as existing lines updated on behalf of the emp, sal columns.

  Start timestamp to join date and the current flag to flag column.

  first ename is smith when empno is 7369.

  I've updated ename

  Update scdtype1_source set ENAME = 'AA' where EMPNO = 7369

  After running, I don't have to change the flag o to 1 in the target table.

  SCD type1 mainly used for replaces old values., but I add flag column in the target table for easy identification. () column in which I updated)

  I have not any change in the flag.

  Please help me,

  Thnaks in advance,

  A.Kavya

  

  Hello

  I'm not sure what you're trying to do will work. In the case of Type 1 SCD are you still using the dimension of change slow ikm oracle? The flag is updated to 0 in the case of CDs because of the code in step 2 "' Historize old rows where CPCs flag is set to '0'.» In your case, that no new record is added in the case of an update so the IKM does not change the flag.

  I'm not sure of your requirement either. You say you want to identify the column in which you have updated. How is an indicator will tell you what the column value has been changed? Maybe you want to know which line you have updated? If this is the case and lets say you set the flag to 1 in case of update what happens when another updated same record?

  I would suggest you add an updated time stamp column to your table and set of sysdate but in properties check only the update and will not insert. What that would do is update your record, it will fill the current date and time so that you can identify that this record has been updated.

  If you want to always go for the flag, then follow the same approach but in your mapping, change the value of the flag to 1 instead of 0 and new check only the update box (uncheck the box insert). In your create table statement the default value of this option to 0. So when the new line is inserted database is automatically assigned the flag to 0 and where there is an update, ODI it will change to 1.

  Thank you

  Ajay

• The Master Table column updated based on the sum of column Table detail

  
  With the help of JDev 11.1.1.6.

  I have a master-detail table based on a link to BC.

  The main table has a column that displays an InputText or an OutputText, based on the value in another column.

  If the InputText is displayed, the user can enter a value and the database will be updated with that value.

  If the OutputText is displayed, it must be a sum of a column in the secondary table.  Also, this value will be written in the database.

  Question:

  How can I fill the OutputText in the main table with the sum of the values in a column in the secondary table?

  The detail table column is a manually entered InputText field.

  Thank you.

  Create a spike in the main table and write in its expression as follows - DetailVoAccessorName.sum ("ColumnName");

  This will calculate the sum of column table detail and then you can set the value of the transient attribute to attribute DB on backup operation

  Ashish

• Column update

  Hello

  I have a problem that cannot be resolved.

  I have a table called BEFORE:

  CREATE TABLE 'BEFORE' ('ID' VARCHAR2(6 BYTE), NUMBER (2.0)) 'LENGTH '.

  /

  REM, INSERTING BEFORE

  TOGETHER TO DEFINE

  Insert FRONT (ID, LENGTH) values ('125716 ', ' 11');

  Insert FRONT (ID, LENGTH) values ('125717 ', ' 11');

  Insert FRONT (ID, LENGTH) values ('126816 ', ' 11');

  Insert FRONT (ID, LENGTH) values ('126817 ', ' 11');

  Insert FRONT (ID, LENGTH) values ('126818 ', ' 7');

  Insert FRONT (ID, LENGTH) values ('126819 ', ' 7');

  Insert FRONT (ID, LENGTH) values ('127518 ', ' 7');

  Insert FRONT (ID, LENGTH) values ('127519 ', ' 7');

  Insert FRONT (ID, LENGTH) values ('127520 ', ' 5');

  Insert FRONT (ID, LENGTH) values ('127521 ', ' 5');

  Insert FRONT (ID, LENGTH) values ('128020 ', ' 5');

  Insert FRONT (ID, LENGTH) values ('128021 ', ' 5');

  Insert FRONT (ID, LENGTH) values ('128022 ', ' 4');

  Insert FRONT (ID, LENGTH) values ('128023 ', ' 4');

  Insert FRONT (ID, LENGTH) values ('128422 ', ' 4');

  Insert FRONT (ID, LENGTH) values ('128423 ', ' 4');

  I have to update the ID column, so that if two lines with the same number in the column 'Length'and as the first four digits of the 'ID' column are identical, then the two last digits in columns 'ID' must be identical too.

  I have to get this result:

  CREATE TABLE 'AFTER' ('ID' VARCHAR2(6 BYTE), NUMBER (2.0)) 'LENGTH '.

  /

  REM in AFTER INSERTION

  TOGETHER TO DEFINE

  Insert AFTER (ID, LENGTH) values ('125716 ', ' 11');

  Insert AFTER (ID, LENGTH) values ('125716 ', ' 11');

  Insert AFTER (ID, LENGTH) values ('126817 ', ' 11');

  Insert AFTER (ID, LENGTH) values ('126817 ', ' 11');

  Insert AFTER (ID, LENGTH) values ('126818 ', ' 7');

  Insert AFTER (ID, LENGTH) values ('126818 ', ' 7');

  Insert AFTER (ID, LENGTH) values ('127519 ', ' 7');

  Insert AFTER (ID, LENGTH) values ('127519 ', ' 7');

  Insert AFTER (ID, LENGTH) values ('127520 ', ' 5');

  Insert AFTER (ID, LENGTH) values ('127520 ', ' 5');

  Insert AFTER (ID, LENGTH) values ('128021 ', ' 5');

  Insert AFTER (ID, LENGTH) values ('128021 ', ' 5');

  Insert AFTER (ID, LENGTH) values ('128022 ', ' 4');

  Insert AFTER (ID, LENGTH) values ('128022 ', ' 4');

  Insert AFTER (ID, LENGTH) values ('128423 ', ' 4');

  Insert AFTER (ID, LENGTH) values ('128423 ', ' 4');

  Thank you if you have an idea.

  Lalit Kumar B wrote:

  Then, simply MIN won't work. OP wants the last two digits to be ascending after each set of 2 rows.

  I see, I did not increase the share of the last two digits:

  INSERT

  IN AFTER

  SELECT SUBSTR (ID, 1, 4). (MIN (SUBSTR(ID,5)) OVER() +)

  DENSE_RANK() OVER (ORDER OF LENGTH DESC, SUBSTR (ID, 1, 4))) - 1 ID,.

  LENGTH

  FRONT

  /

  16 lines were created.

  SQL > SELECT *.
  2 ACCORDING
  3 ORDER BY DESC LENGTH,
  4             ID
  5.

  THE ID LENGTH
  ------ ----------
  125716 11
  125716 11
  126817 11
  126817 11
  126818 7
  126818 7
  127519 7
  127519 7
  5 127520
  5 127520
  128021 5

  THE ID LENGTH
  ------ ----------
  128021 5
  128022 4
  128022 4
  128423 4
  128423 4

  16 selected lines.

  SQL >

  SY.

• Date column updated with trigger update of the geometry

  Hello
  
  I have a column named "GEO_MOD_DATE" which is supposed to be updated every time that my geometry column is updated. My trigger code is below. Currently the trigger is fired when other than my column of geometry columns are updated. I searched the forum and many other sites. As far as I know, my trigger is written correctly. But I can't understand why it is taken when the columns except my geometry column are updated. You have any ideas? Thanks in advance! Note that the trigger fires when I update the geometry column. However, it is also SHOOTING when the other columns are updated. I have also tried adding 'OF GDO_GEOMETRY' from "BEFORE UPDATE" paragraph and which did not help.
  
  Jeff
  
  Triggering factor:
  
  CREATE OR REPLACE TRIGGER OPER_ZONE_VALVE_GEO_BU_T
  BEFORE THE UPDATE
  ON OPER_ZONE_VALVE
  REFERRING AGAIN AS NINE OLD AND OLD
  FOR EACH LINE
  DECLARE
  v_foo NUMBER (9);
  BEGIN
  -Comment from below if you do not want to assign GEO_MOD_DATE
  BEGIN
  IF the UPDATE ("GDO_GEOMETRY") and: NEW.gdo_geometry IS NOT NULL
  THEN
  : NEW.geo_mod_date: = SYSDATE;
  ON THE OTHER
  : NEW.geo_mod_date: = NULL;
  END IF;
  EXCEPTION
  WHILE OTHERS
  THEN
  raise_application_error
  (- 20013,
  "Cannot be autoassign field OPER ZONE VALVE".
  );
  END;
  
  /*****/
  NULL;
  EXCEPTION
  WHILE OTHERS
  THEN
  raise_application_error
  (- 20013,
  'error in executing BEFORE INSERT TRIGGER OPER_ZONE_VALVE_GEO_BU_T;'
  );
  END;
  /
  
  --------------------------------------------
  And here is my table structure:
  
  CREATE TABLE OPER_ZONE_VALVE
  (
  MSLINK NUMBER (38),
  VALVE_NUMBER VARCHAR2 (8 BYTE),
  ECM_NUMBER VARCHAR2 (8 BYTE),
  VALVE_SIZE VARCHAR2 (4 BYTE),
  CONNECTION_TYPE VARCHAR2 (100 BYTE),
  HEADER_STREET VARCHAR2 (100 BYTE),
  HEADER_FEET VARCHAR2 (8 BYTE),
  HEADER_DIR VARCHAR2 (4 BYTE),
  HEADER_PROP VARCHAR2 (4 BYTE),
  SUB_STREET VARCHAR2 (100 BYTE),
  SUB_FEET VARCHAR2 (8 BYTE),
  SUB_DIR VARCHAR2 (4 BYTE),
  SUB_PROP VARCHAR2 (4 BYTE),
  PLAT_MAP_NUMBER VARCHAR2 (4 BYTE),
  USER DEFAULT CREATED_BY VARCHAR2 (50 BYTES)
  CREATED_DATE DATE DEFAULT SYSDATE,
  MODIFIED_BY VARCHAR2 (50 BYTE),
  MODIFIED_DATE DATE,
  DATE OF GEO_MOD_DATE,
  NOTE VARCHAR2 (40 BYTE),
  MDSYS GDO_GEOMETRY. SDO_GEOMETRY
  )

  I read that FORWARD updated TO does not work with data as CLOB types

  Yes. But what does this mean for the SDO_GEOMETRY objects is that you cannot directly use a WHEN clause in the definition of the trigger and you cannot compare two geometries directly via If (: old.geom =: new.geom) then.

  So, in your trigger update of your business rules conditions require the trigger to implement are not enough.

  Try to check if the geometry column has actually changed.

  DROP TABLE oper_zone_valve;
  
  CREATE TABLE oper_zone_valve
  (
    mslink            NUMBER (38),
    valve_number      VARCHAR2 (8 BYTE),
    ecm_number        VARCHAR2 (8 BYTE),
    valve_size        VARCHAR2 (4 BYTE),
    connection_type   VARCHAR2 (100 BYTE),
    header_street     VARCHAR2 (100 BYTE),
    header_feet       VARCHAR2 (8 BYTE),
    header_dir        VARCHAR2 (4 BYTE),
    header_prop       VARCHAR2 (4 BYTE),
    sub_street        VARCHAR2 (100 BYTE),
    sub_feet          VARCHAR2 (8 BYTE),
    sub_dir           VARCHAR2 (4 BYTE),
    sub_prop          VARCHAR2 (4 BYTE),
    plat_map_number   VARCHAR2 (4 BYTE),
    created_by        VARCHAR2 (50 BYTE) DEFAULT USER,
    created_date      DATE DEFAULT SYSDATE,
    modified_by       VARCHAR2 (50 BYTE),
    modified_date     DATE,
    geo_mod_date      DATE,
    remark            VARCHAR2 (40 BYTE),
    gdo_geometry      MDSYS.sdo_geometry
  );
  
  CREATE OR REPLACE TRIGGER air_valve_geo_bu_t
    BEFORE UPDATE OF gdo_geometry
    ON oper_zone_valve
    REFERENCING NEW AS new OLD AS old
    FOR EACH ROW
  BEGIN
     IF ( NOT UPDATING('GDO_GEOMETRY') ) THEN
      RETURN;
    END IF;
    IF ( :old.gdo_geometry is not null and :new.gdo_geometry IS not NULL ) THEN
      -- Check if geometry has changed internally
       IF ( sdo_geom.relate(:old.gdo_geometry,'DETERMINE',:new.gdo_geometry,0.005) != 'EQUAL' ) Then
         :new.geo_mod_date := SYSDATE;
       End If;
    ELSIF ( ( :old.gdo_geometry is null and :new.gdo_geometry IS NOT NULL ) OR
         ( :old.gdo_geometry is not null and :new.gdo_geometry IS NULL ) ) THEN
       :new.geo_mod_date := SYSDATE;
    ELSE
      :new.geo_mod_date := NULL;
    END IF;
  END;
  /
  SHOW ERRORS
  

  Some tests

  SET NULL NULL
  INSERT INTO oper_zone_valve ( remark, gdo_geometry) VALUES ('remark', MDSYS.sdo_geometry(2001,NULL,SDO_POINT_TYPE(0,0,0), NULL,NULL)); COMMIT;
  1 rows inserted.
  commited.
  SELECT remark,  to_char(geo_mod_date,'YYYY-MM-DD HH24:MI:SS') as geo_mod_date, gdo_geometry FROM oper_zone_valve;
  REMARK GEO_MOD_DATE        GDO_GEOMETRY
  ------ ------------------- -------------------------------------------------------
  remark NULL                SDO_GEOMETRY(2001,NULL,SDO_POINT_TYPE(0,0,0),NULL,NULL)
  
  update oper_zone_valve set gdo_geometry = sdo_geometry(2001, NULL,SDO_POINT_TYPE( 0,0,0),NULL,NULL) where remark = 'remark'; COMMIT;
  1 rows updated.
  commited.
  SELECT remark,  to_char(geo_mod_date,'YYYY-MM-DD HH24:MI:SS') as geo_mod_date, gdo_geometry FROM oper_zone_valve;
  REMARK GEO_MOD_DATE        GDO_GEOMETRY
  ------ ------------------- -------------------------------------------------------
  remark NULL                SDO_GEOMETRY(2001,NULL,SDO_POINT_TYPE(0,0,0),NULL,NULL)
  
  update oper_zone_valve set gdo_geometry = sdo_geometry(2001, NULL,SDO_POINT_TYPE(10,0,0),NULL,NULL) where remark = 'remark'; COMMIT;
  1 rows updated.
  commited.
  SELECT remark,  to_char(geo_mod_date,'YYYY-MM-DD HH24:MI:SS') as geo_mod_date, gdo_geometry FROM oper_zone_valve;
  REMARK GEO_MOD_DATE        GDO_GEOMETRY
  ------ ------------------- --------------------------------------------------------
  remark 2012-10-16 09:24:53 SDO_GEOMETRY(2001,NULL,SDO_POINT_TYPE(10,0,0),NULL,NULL)
  
  execute dbms_lock.sleep(5);
  anonymous block completed
  update oper_zone_valve set gdo_geometry = NULL  where remark = 'remark'; COMMIT;
  1 rows updated.
  commited.
  SELECT remark,  to_char(geo_mod_date,'YYYY-MM-DD HH24:MI:SS') as geo_mod_date, gdo_geometry FROM oper_zone_valve;
  REMARK GEO_MOD_DATE        GDO_GEOMETRY
  ------ ------------------- --------------------------------------------------------
  remark 2012-10-16 09:24:58 NULL
  

  This seems to be the answer you are looking for or puts you on a path to achieve the correct execution of your business rules in the trigger.

  Shows please!

  concerning
  Simon

  Published by: Simon Greener on 17 October 2012 15:39, fixed a display issue does not

• SmartView VBA - can't set "Width of the column updated"

  Hi people,
  
  I am updating a series of workbooks based on the API VBA Essbase v9 on SmartView 11.1.2.1 and touched a problem - I am new to VBA SV commands, and I hope that it is something obvious that I'm missing.
  
  One of the functions which is essential to ensure recovery of data does not affect the layout of spreadsheets is the parameter 'Adjust the column widths', which, in the VBA Essbase API, could be managed by using the EssVSetSheetOption function, but neither HypSetGlobalOption nor HypSetSheetOption appear to have such a capability. The only option I have is to run get followed by some reformatting of the column widths, which seems rather ineligant for me.
  
  Any ideas?
  
  Thank you very much
  
  Dave

  research in the .bas file (and what needs to be included in your project that I find HSV_ADJUSTCOLUMNWIDTH in the definitions of the index count options to use for HypGetOption/HypSetOption

  Note, I'm looking at the 11.1.2.2 version, your mileage may vary depending on your version

• question on multi column update

  Hi all
  
  I'm trying a few updates on a column of several updates.
  
  This is the update that I am trying to run it.
  
  Can some pls check if what I try to do is correct or not.
  You want to know if I do no redundant checks on the data during the update of the table.
  
  
  UPDATE
  land_pp_sub_feature_temp_1 T4
  SET
  (
  esp_message_id
  esp_message_id_dt
  event_bus_channel
  esp_event_type
  event_source_system
  event_create_dt
  profile_id
  event_receive_id
  valid_to_dt
  feature_code
  transaction_dt
  ) =
  (
  SELECT
  esp_message_id
  esp_message_id_dt
  event_bus_channel
  esp_event_type
  event_source_system
  event_create_dt
  profile_id
  event_receive_id
  valid_to_dt
  feature_code
  transaction_dt
  sub_id
  msisdn
  feature_code
  Of
  (
  SELECT
  T1.esp_message_id
  T1.esp_message_id_dt
  T1.event_bus_channel
  T1.esp_event_type
  T1.event_source_system
  T1.event_create_dt
  T1.profile_id
  T1.event_receive_id
  T1.valid_to_dt
  T1.feature_code
  T1.transaction_dt
  T1.sub_id
  T1.msisdn
  T1.feature_code
  OF hold_pp_sub_feature_expire T1, land_pp_sub_feature_temp_1 T2
  WHERE t1.sub_id = t2.sub_id
  AND t1.msisdn = t2.msisdn
  AND t1.feature_code = t2.feature_code;
  ) T3
  WHERE t3.sub_id = t4.sub_id
  AND t3.msisdn = t4.msisdn
  AND t3.feature_code = t4.feature_code
  )
  WHERE (msisdn, sub_id, feature_code)
  IN (SELECT sub_id, msisdn, feature_code FROM hold_pp_sub_feature_expire);
  
  Thank you
  MK.

  May be you can simplify it like this

   update land_pp_sub_feature_temp_1 t2
      set
       (
              esp_message_id
            , esp_message_id_dt
            , event_bus_channel
            , esp_event_type
            , event_source_system
            , event_create_dt
            , profile_id
            , event_receive_id
            , valid_to_dt
            , feature_code
            , transaction_dt
       ) =
       (
              select  t1.esp_message_id
                 , t1.esp_message_id_dt
                 , t1.event_bus_channel
                 , t1.esp_event_type
                 , t1.event_source_system
                 , t1.event_create_dt
                 , t1.profile_id
                 , t1.event_receive_id
                 , t1.valid_to_dt
                 , t1.feature_code
                 , t1.transaction_dt
                 , t1.sub_id
                 , t1.msisdn
                 , t1.feature_code
                 from hold_pp_sub_feature_expire t1
                where t1.sub_id       = t2.sub_id
                  and t1.msisdn       = t2.msisdn
                  and t1.feature_code = t2.feature_code
       )
    where
   exists (
            select null
              from hold_pp_sub_feature_expire t1
             where t1.sub_id       = t2.sub_id
               and t1.msisdn       = t2.msisdn
               and t1.feature_code    = t2.feature_code
       )
            
  

  Just make sure that the combination of (SUB_ID, MSISDN, feature_code) is unique in the table two