Big history table based on SCD2

Ola,
Has anyone of you ever experienced difficulties running millions of records through a SCD2 mapping? I created a couple of history tables and created mappings to fill them, based on SCD2-method. When I tested the mappings with a subset of data (10.000 - 40-000 rows), it all went OK.
Now I test the mapping with the full data set and three of the most large tables fail. The largest table is 23.7 million records. The smallest of the failing ones is 17.8 million. I get an error like this:
ORA-06502: PL/SQL: numeric or value error: character string buffer too small ORA-06512: at "OWNER_OWR.WB_RT_MAPAUDIT", line 1762 ORA-06512: at "OWNER_OWR.WB_RT_MAPAUDIT", line 2651 ORA-06512: at "OWNER_LCN.E_ASL_SERVICE_HIST_LASH", line 3159 ORA-01555: snapshot too old: rollback segment number 2 with name "_SYSSMU2$" too small ORA-06512: at "OWNER_LCN.E_ASL_SERVICE_HIST_LASH", line 3680 ORA-06512: at "OWNER_LCN.E_ASL_SERVICE_HIST_LASH", line 4229 ORA-06512: at line 1
Only thing I changed is the dataset. Nothing else...
Does the 'snapshot too old' message mean that it takes too much time to finish the mapping process?
Anyone any clue?
Kind regards,
Moscowic

Though I havent faced the error while using OWB, saw this error on other occasaions when the the transaction was pretty large or if the UNDO segment was too small and couldnt hold the rollback information of a transaction.
Check a couple of things...
1. if you are running mapping it in a SET based mode with large data, u r likely to hit this error.
2. Ask your DBA to lookinto the UNDO segment/tablesapce, if in case any fragmentation issues.
Hope this helps.
Nat

Similar Messages

Moving Data from Normal table to History tables

Hi All,
I'm in the process of moving data from normal tables to
History tables.
It can be some sort of a procedure which should be a cron job running at night.
My aim is to move data say 1.5 yrs or 2yrs old data to History tables.
What aspects i need to check when moving data. And how can i write a procedure for this requirement.
The schema is same in both the normal table and history table.
It has to be a procedure based on particular field RCRE_DT.
If the rcre_dt is above 2 yrs the data needs to be moved to HIS_<table>.
I have to insert record in to HIS_table and simultaneously delete record from the normal table.
This is in Production system and the tables are quite big.
Pls do find enclosed the attached sample schema for Normal table and HIS_<table>.
If i want to automate this script as a Cron job for similarly other History tables
how am i to do it in a single procedure assuming the procedure for moving the data is the same procedure.
Thanks for ur help in advance.
SQL> DESC PXM_FLT;
Name Null? Type
RCRE_USER_ID NOT NULL VARCHAR2(15)
RCRE_DT NOT NULL DATE
LCHG_USER_ID VARCHAR2(15)
LCHG_DT DATE
AIRLINE_CD NOT NULL VARCHAR2(5)
REF_ID NOT NULL VARCHAR2(12)
BATCH_DT NOT NULL DATE
CPY_NO NOT NULL NUMBER(2)
ACCRUAL_STATUS NOT NULL VARCHAR2(1)
FLT_DT NOT NULL DATE
OPERATING_CARRIER_CD NOT NULL VARCHAR2(3)
OPERATING_FLT_NO NOT NULL NUMBER(4)
MKTING_CARRIER_CD VARCHAR2(3)
MKTING_FLT_NO NUMBER(4)
BOARD_PT NOT NULL VARCHAR2(5)
OFF_PT NOT NULL VARCHAR2(5)
AIR_CD_SHARE_IND VARCHAR2(1)
UPLOAD_ERR_CD VARCHAR2(5)
MID_PT1 VARCHAR2(5)
MID_PT2 VARCHAR2(5)
MID_PT3 VARCHAR2(5)
MID_PT4 VARCHAR2(5)
MID_PT5 VARCHAR2(5)
PAX_TYPE VARCHAR2(3)
PAY_PRINCIPLE VARCHAR2(1)
SQL> DESC HIS_PXM_FLT;
Name Null? Type
RCRE_USER_ID NOT NULL VARCHAR2(15)
RCRE_DT NOT NULL DATE
LCHG_USER_ID VARCHAR2(15)
LCHG_DT DATE
AIRLINE_CD NOT NULL VARCHAR2(5)
REF_ID NOT NULL VARCHAR2(12)
BATCH_DT NOT NULL DATE
CPY_NO NOT NULL NUMBER(2)
ACCRUAL_STATUS NOT NULL VARCHAR2(1)
FLT_DT NOT NULL DATE
OPERATING_CARRIER_CD NOT NULL VARCHAR2(3)
OPERATING_FLT_NO NOT NULL NUMBER(4)
MKTING_CARRIER_CD VARCHAR2(3)
MKTING_FLT_NO NUMBER(4)
BOARD_PT NOT NULL VARCHAR2(5)
OFF_PT NOT NULL VARCHAR2(5)
AIR_CD_SHARE_IND VARCHAR2(1)
UPLOAD_ERR_CD VARCHAR2(5)
MID_PT1 VARCHAR2(5)
MID_PT2 VARCHAR2(5)
MID_PT3 VARCHAR2(5)
MID_PT4 VARCHAR2(5)
MID_PT5 VARCHAR2(5)
PAX_TYPE VARCHAR2(3)
PAY_PRINCIPLE VARCHAR2(1)

Hi All,
Thanks for ur valuable suggestion.But can u explain me bit more on this as i'm still confused about switching between partitoned tables and temporary table.Suppose if i have a table called PXM_FLT and an correspoding similar table named HIS_PXM_FLT.How can i do the partitioning shd i do the partitioning on the normal table or HIS_PXM_FLT.i do have a date field for me to partition based on range.Can u pls explain why shd i again create a temp.table.What's the purpose.Now the application is designed in such a way that old records have to be moved to HIS_PXM_FLT.can u pls highlight more on this.Your suggestions are greatly appreciated.As i'm relatively new to this partitioning technique i'm bit confused on how it works.But i came to understand Partitioning is a better operation than the normal insert or delte as it is more data intensive as millions of record need to be moved.Thanks for feedback and ur precious time.

Updating one table based on another

Ok, that sounds a little confusing, but this is what I want to do:
Table A is my source table and is a temp table and will be reloaded daily.
Table B is my target table.
What I want to do is update Table B using a subset of Table A based on a date within Table B. A simple psuedo code would look like:
update Table B
insert * from Table A where Table B.datefield <> Table A.datefield.
Both tables contain the same data elements, so the structure is an exact 1 to 1 match.
Table B was initially created from Table A using the copy command to do an intial load. From this point forward, I just need to update Table B with new data from Table A, which has data loaded into it throughout the day. My idea is to grab just the data that isn't in Table B once a day.
I'm open to all suggestions or ideas............
Doug

By cracky, I think that merge thing is just what I'm looking for.
In theory the data in Table A will contain the data in Table B, plus newer data. The older data in Table A is stagnant, meaning that that table is just appended to. The problem is, Table A is periodically trimmed by the application. Table B is to be used for compiling all the data and reporting trend and historical data.
More, specifically, Table A is the History table in HP Openview. Table B will be a table in HP Performance Insight. Since theses are 2 totally seperate databases, I will need to copy over the OpenView History table to a temp table in PI nightly, then cull out the records that were added during that day and place those and only those into the PI table for reporting.
I'll give the merge command a try. I think it will do exactly what i need.
Thanks,
Doug

Update table based on ID reference comparing each column and value

Hi,
Through UI user can update any information and click save. In backend i will be receiving only ID as reference . Based on the ID value i should update multiple tables wherever i have the reference tables. Here i will not get the values or column name updated.
I just get a ID to find which row is updated.
I should do a comparision now.
I have a history table where i get the row before updation. and i get a row after updation from current data i.e. from ssdb_Railroad[This will be updated by java after updation we will get the ID inorder to update remaining tables]. Now i need to compare
both the table column values and know which column is updated and after getting the value i should update reference tables.
Below is the structure for comparision table , History table and Main table.
Create table Comparision_History (
ID int IDENTITY(1,1) PRIMARY KEY,
InsertUser_UpdateUser varchar(50),
Old_ColVal varchar(100),
New_ColVal varchar(100),
TableName varchar(100),
[Action] varchar(50),
InsertDate_UpdateDate Datetime NOT NULL DEFAULT getdate()
CREATE TABLE SSDB_Railroad_History (
ID int IDENTITY PRIMARY KEY,
SCAC varchar(4) ,
Name varchar(50) NOT NULL,
RailroadID int NOT NULL UNIQUE,
RC2Key
varchar(50),
NOTES varchar(1000),
[Action] varchar(50),
InsertDate_UpdateDate Datetime NOT NULL DEFAULT getdate(),
InsertUser_UpdateUser varchar(50),
CREATE TABLE SSDB_Railroad (
ID int IDENTITY(1,1)PRIMARY KEY,
SCAC varchar(4) UNIQUE,
Name varchar(50) NOT NULL,
RailroadID int NOT NULL UNIQUE,
RC2Key
varchar(50),
InsertDate Datetime NOT NULL DEFAULT getdate(),
UpdateDate Datetime,
InsertUser varchar(50),
UpdateUser varchar(50)
Here SSDB_Railroad table and History table needs to be compared and get a updated value and it should be inserted in a comparision table. as well as it should be updated in reference tables.
Please help me with the code.
Thanks in Advance
Deepa

Hi Deepa_Deepu,
According to your description, personally, the merge function can meet your requirements. Usually, we can synchronize two tables by updating or inserting rows in a target table based on differences found in the source tables. Just as your description, when
the value of ID and InsertDate_UpdateDate in the source table matches a value, update them in the target table. When the values does not match, the source row is inserted into the target table.
For more information, see: using MERGE to perform UPDATE and INSERT operations on a target table by using a derived source table.
http://msdn.microsoft.com/zh-cn/library/bb510625.aspx
Regards,
Sofiya Li
Sofiya Li
TechNet Community Support

History tables on 8.8

Guys,
We've got various add-on solutions, and reports, that examine the history tables for various B1 objects
(documents, Business Partners, etc).
Up until 8.8, the history tables always stored the current revision of a record as well as the historical
revisions. In 8.8, it seems that the current revision is no longer stored in the history tables - the highest
revision stored is the one previous to the current state.
This has quite an impact on our add-ons and reports - does anyone know if this behaviour is by design,or a bug in 8.8?

For example, if your create a Sales Order under 2005 or 2007, the ordr table is populated with the sales order and its values. Also a record is created in the adoc table with a loginstanc of 1, with those same (current values). Now edit the SO, (by altering the comments/remarks field for instance). What happens now is that the record in the ordr table reflects your change, as you would expect. In the adoc table, a new record with a loginstanc of 2 is created, containing these current values, so you now have two records in adoc, the first being the original data, the second being the current data.
Now, do the same in 8.8 (I have PL06 installed). When you create the SO, the order table is populated as you would expect. However, no record is created in adoc at all. When you edit the SO, a record is created in adoc with a loginstanc of 1 which contains the original data.
Now, we have add ons and reports that work on the historical data (not just for documents, for other objects too) that rely on thebehaviour shown in 2005/2007 as it means we don't have to link back to the other tables to get the current data - which is a big bonus when working with adoc, otherwise you have to join dependant on the value of objtype.
If this has changed by design in 8.8 it's going to cause us quite a bit of pain.
I can't find any info on this in the 8.8 documents, nor in the list of changes in the database/SDK for 8.8.
I just want to know if this is by design, or it's a bug in the ramp-up versions of 8.8.
I'm also not sure what's going to happen when we upgrade a 2007 DB to 8.8 - will it strip out the current revisions from teh history tables? I might try that today.

How to insert past record after updating the master table in history table through store PROC

Master Table
Party Status
A Active
B Inactive
C Active
D Inactive
Duplicate Table
Party Status
A Active
B Active
C Active
D Inactive
Updated Master Table
Party Status
A Active
B Active
C Active
D Inactive
Party History Table
B Inactive
I have two table one master and another duplicate I need to update master table based on duplicate table insert the record which updated into Party history table as shown above. need help to write store proc.

Check MERGE syntax in BOL (example D). There should be a sample with output, e.g.
insert into PartyHistory (Party, [Status])
select Party, [Status] FROM
(MERGE Master M using Duplicate D on M.[Party]=D.[Party]AND M.[Status]<>D.[Status]
WHEN MATCHED THEN UPDATE
SET [Status] = D.[Status]
OUTPUT Deleted.[Party], Deleted.[Status], $Action) AS Changes (Party, [Status], Action) WHERE Action = 'UPDATE'
For every expert, there is an equal and opposite expert. - Becker's Law
My blog
My TechNet articles

Change background color of a table based on table cell values quickly.

Is there a more effecient way to change the background color of a table based on the cell values of the table?
I can do this cell-by-cell using the property node, but it is very slow. It helps a little by deferring panel update, but it is still slow. Is there a better way?
Thanks.
Ian

If you want to color each cell according to the value, you need to do it one cell at a time and the above answers don't apply.
You can dramatically speed up things by defering front panel updates for the duration of the operation.
(sample code can be seen in the top picture here)
EDIT: I noticed you tried this already. Can you show us your code?
How many cells need to be colored on average? If there are not that many, first write the background color using -2,-2, the only write the few cells to be colored, skipping the rest.
How big is the table? If it is much bigger than the table indicator, all you need is to color the visible parts whenever the data or the scroll position changes.
LabVIEW Champion . Do more with less code and in less time .

How to populate data in the same table based on different links/buttons

Hi
I'm using jdeveloper 11.1.4. I have a use case in which i need to populate data in the same table based on click of different links.
Can anyone please suggest how can this be achieved.
Thanks

I have a use case in which i need to populate data in the same table based on click of different linksDo you mean that you need to edit existing rows ?
What format do you have the date in - table / form ?

Problem with parameters and non table based items

Hi,
I have a situation where I have some not table based items on a page (shuttles and select lists). It is parameters that the end user inputs. I have another page with the same items just here they are Display Only and also not table based. When the user press a button I branch from page 1 to page 2 and set the items on page 2. My problem is that the input the user keys in are not copied to the items on page 2. I can see the data in the session state.
I assume that this is a silly question and I have just set Source Used or Source Type wrong one one of the items but I can not get it working. Can anybody help?
It works fine if I use a Application Item and creates a process on page 1 where I copy the item value from the session state but that should be needed right?
Regards Pete

Thanks Scott,
I think I got it working. If I remove all the colons from the data it works. I select several entries in a Shuttle and the result is a colon delimited string of values. If I remove these in a process it works fine. Is there a way of escaping the output from a Shuttle?
Regards Pete

How can i update rows in a table based on a match from a select query

Hello
How can i update rows in a table based on a match from a select query fron two other tables with a update using sqlplus ?
Thanks Glenn
table1
attribute1 varchar2 (10)
attribute2 varchar2 (10)
processed varchar2 (10)
table2
attribute1 varchar2 (10)
table3
attribute2 varchar2 (10)
An example:
set table1.processed = "Y"
where (table1.attribute1 = table2.attribute1)
and (table1.attribute2 = table3.attribute2)

Hi,
Etbin wrote:
Hi, Frank
taking nulls into account, what if some attributes are null ;) then the query should look like
NOT TESTED !
update table1 t1
set processed = 'Y'
where exists(select null
from table2
where lnnvl(attribute1 != t1.attribute1)
and exists(select null
from table3
where lnnvl(attribute2 != t1.attribute2)
and processed != 'Y'Regards
EtbinYes, you could do that. OP specifically requested something else:
wgdoig wrote:
set table1.processed = "Y"
where (table1.attribute1 = table2.attribute1)
and (table1.attribute2 = table3.attribute2)This WHERE clause won't be TRUE if any of the 4 attribute columns are NULL. It's debatable about what should be done when those columns are NULL.
But there is no argument about what needs to be done when processed is NULL.
OP didn't specifically say that the UPDATEshould or shouldn't be done on rows where processed was already 'Y'. You (quite rightly) introduced a condition that would prevent redo from being generated and triggers from firing unnecessarily; I'm just saying that we have to be careful that the same condition doesn't keep the row from being UPDATEd when it is necessary.

Hide multiple rows in a dynamic table based on the row value.

Hi,
I need to hide multiple rows in a dynamic table based on the specific value of that row.
I cant find the right expression to do that.
please help

Go to the Row Properties, and in the Visibility tab, you have "Show or hide based on an expression". You can use this to write an expression that resolves to true if the row should be hidden, false otherwise.
Additionally, in the Matrix properties you should take a look at the filters section, perhaps you can achieve what you wish to achieve through there by removing the unnecessary rows instead of just hiding them.
It's only so much I can help you with the limited information. If you require further help, please provide us with more information such as what data are you displaying, what's the criteria to hiding rows, etc...
Regards
Andrew Borg Cardona

How to get a block of data from internal table based on a criteria

Hi all,
I have some records in the internal table t_int1. I want to retrieve some set records from that table and put them all in some other table, based on a single field which is not a key. Can i use READ statement to achieve this.
Could you please let me know any simple way of doin this.
Regards,
Vishnu

I have some records in the internal table t_int1. I want to retrieve some set records from that table and put them all in some other table, based on a single field which is not a key. Can i use READ statement to achieve this.
Could you please let me know any simple way of doin this.
Answer :
data:
itab2 like standard table of t_int1 with header line.
Loop at t_int1.
read table t_int1 with index 1.
check ur condition----
check each line and insert it ---work out the syntax for this
IF t_int1-xyz = data1
append line of t_int1 to itab2. or try insert
cnt = cnt + 1 -
u will get no. of records added to next itab.
else.
cnt1 = cnt1 + 1 -
u will get no. of records not added to next itab.
endif.
end loop.

Error updating tables based on schema

Hello,
I'm trying to update a table based on the next schema:
<?xml version="1.0" encoding="UTF-8"?>
<xs:schema xmlns:xdb="http://xmlns.oracle.com/xdb" xmlns:xs="http://www.w3.org/2001/XMLSchema" elementFormDefault="qualified" attributeFormDefault="unqualified">
     <xs:element name="HILO" xdb:SQLType="HILO_TYPE" xdb:defaultTable="HILO_TABLE">
          <xs:complexType>
               <xs:sequence>
                    <xs:element name="MENSAJE" maxOccurs="unbounded" xdb:SQLType="MENSAJE_TYPE" xdb:maintainOrder="false">
                         <xs:complexType>
                              <xs:sequence>
                                   <xs:element name="FECHA" type="xs:date" xdb:SQLType="DATE"/>
                                   <xs:element name="TITULO" xdb:SQLType="VARCHAR2">
                                        <xs:simpleType>
                                             <xs:restriction base="xs:string">
                                                  <xs:maxLength value="200"/>
                                             </xs:restriction>
                                        </xs:simpleType>
                                   </xs:element>
                                   <xs:element name="CUERPO" type="xs:string" xdb:SQLType="VARCHAR2"/>
                                   <xs:element name="DNI_CREADO_POR" xdb:SQLType="VARCHAR2">
                                        <xs:simpleType>
                                             <xs:restriction base="xs:string">
                                                  <xs:maxLength value="9"/>
                                             </xs:restriction>
                                        </xs:simpleType>
                                   </xs:element>
                                   <xs:element name="ASIGNATURA" xdb:SQLType="VARCHAR2">
                                        <xs:simpleType>
                                             <xs:restriction base="xs:string">
                                                  <xs:maxLength value="3"/>
                                             </xs:restriction>
                                        </xs:simpleType>
                                   </xs:element>
<xs:element name="APROBADO" xdb:SQLType="VARCHAR2">
<xs:simpleType>
<xs:restriction base="xs:string">
<xs:enumeration value="APROBADO"/>
<xs:enumeration value="RECHAZADO"/>
<xs:enumeration value="PENDIENTE"/>
</xs:restriction>
</xs:simpleType>
                                   </xs:element>
                              </xs:sequence>
                              <xs:attribute name="NUMERO" type="xs:int" use="required" xdb:SQLType="INTEGER"/>
                         </xs:complexType>
                    </xs:element>
               </xs:sequence>
               <xs:attribute name="CONTADOR" type="xs:int" use="required" xdb:SQLType="INTEGER"/>
          </xs:complexType>
     </xs:element>
</xs:schema>
And the update that i'm trying to do is the next one:
UPDATE hilo_table p
SET value(p) =
updatexml(value(p),'/HILO/MENSAJE/CUERPO/text()', 'Edit' )
where existsnode(VALUE(p), 'HILO[@CONTADOR="1"]') = 1;
ERROR en línea 1:
ORA-01732: operación de manipulación de datos no válida en esta vista
I have updated another tables based on schema and it works fine, any idea?

OK this is your decision. However If you are building a production system that is important to your orgnanization as the product manager I strongly recommend that you either upgrade to a least 9.2.0.3.0 or discontinue using XML DB technology completely
There are probably somewhere in the order of 500 serious XML DB related bugs fixed between 9.2.0.1.0 and 9.2.0.7.0, all of which have been regarded as serious enough to impossible for a customer to develop or deploy a production system. None of these will EVER be fixed or available as patches for 9.2.0.1.0.
The way in which we store the data on the disc is different in 9.2.0.3.0, and while in theory we migrate the on disc format when you upgrade in practice I know that is has never been tested with any significant amount of data present.
Also you really want to take the support situation into account. Image this, it's 3:00am in the morning and your production system fails due one of the bugs that have been fixed in a later release of the product. You cannot find a workaround and you call oracle support for help. They WILL say, sorry there is nothing we can do until you upgrade to 9.2.0.3.0. Now you have to upgrade in hurry and then re-test everything you have done before you can even start working on a fix.
From a business perspective staying on 9.2.0.1.0 and continuing to use the XML DB technology makes no sense. I hope I have made this clear. If you want I will be more than happy to discuss this issue with your development managers.

CM25: Sort standard table based on custom table.

Hi Experts,
I need to sort the CM25 capacity table based on a ZTABLE where we will define the sequence number of each material. So when CM25 is executed the material should get sorted based on sequence number maintained in Ztable. Now the challenge is how to achieve this sorting before we display the capacity table.
anybody has faced such scenerio please help us.
There are few user exits available but not able to find out how to achieve the required functionality.
Regards,
Raghav
Edited by: Julius Bussche on Jul 30, 2009 9:29 AM
In future threads with meaningless subject titles will simply be deleted as you appear to be doing this intentionally

First, let me say that I have no experience with this CM25 table and where/how it is used or displayed, but here's how I look at it:
But what is the real problem?
1. Don't know the user exit / BAdI to be used?
2. Don't know how to do the sorting?
3. .....?
Answer:
1. This will be a matter of trying I would say and reading the documentation of the exit/BAdI.
2. You can do dynamic sorting of an internal table (if it is an internal table that is used before displaying it).
DATA: lv_sort1 TYPE CHAR10 VALUE 'NAME_OF_COLUMN_FROM_ZTABLE',
lv_sort2 TYPE CHAR10 VALUE 'NAME_OF_OTHER_COLUMN_FROM_ZTABLE'.
SORT itab BY (lv_sort1) ASCENDING (lv_sort2) DESCENDING.
Edited by: Micky Oestreich on Jul 30, 2009 7:34 AM

Searching of programs/includes/tables based on string

Hi Experts,
I have a requirement, is there any program or transaction code to search programs/includes/tables based on the fields
(Eg. company code, sales organization, country, currency) in the selection screen or any other method.
Please suggest, if any solution.
Thanks,
GR

Hi,
You will not find a program for this requirement & that too based on a field.
I would suggest if you have the package then SE80 will be the best transaction which will give you all the objects like Tables/programs/FM/Enhancement that come under a package.
Also if you do a whereused list on fields like Company code, Sales organization, Country, you will end up getting a massive list of objects where it is used.
Regards
Abhii

Big history table based on SCD2

Similar Messages

Maybe you are looking for