Lazy loading or indirection for direct to field mapping striing object

Hi,
I need to achieve indirection for String instance variable, currently mapped using direct to field mapping.
This field is a text column in sql server storing large data which is rarely read.
Recently we have been noticing that it is causing lot of IO on our SAN system.
Is there a way indirection can be achieved for this which is currently mapped as direct to field.
I have tried using indirection in transformation mapping. But profiling db indicates above mentioned text column is still part of the select query.
Thanks
Nagesh

Hi Nagesh,
There is no support for lazy loading of DirectToField mappings in TopLink 10.1.3 POJO persistence. Typically, people who have large CLOBs or BLOBs will put them in a separate table to avoid loading them with their owner. If you can't refactor your data model to accomodate this things get more complicated. Do you ever update these large structures are they read only in your application? If they are read only then a very simple approach is to not map the large data column and instead lazily load the data with a report query.
Ideally you would upgrade to TopLink JPA in 11g where lazy loading of Basic mappings is supported. :-)
--Shaun

Similar Messages

Direct to field mapping to a foreign column

I am using toplink workbench to map an attribute to a column in a foriegn table.
Example:
Table Emp {
EmpID,
AddressID, -- (Foriegn Key to the table Address)
StreetName }
Table Address {
AddressID, (Primary Key)
StreetName }
Class Emp{
empId;
addressId;
streetName;
Class Address {
addressId;
streetName;
In the workbench I have the descriptors defined and I am trying to map attribute 'Emp->streetName' to the column 'Address->StreetName'.
I added the the Address table in the advanced properties->Multitable Info of Emp class and also defined the association between the tables.
I mapped the 'streetName' property to the the 'Address->StreetName' using a direct to field mapping.
When I do getter on the streetName of the Emp class I get Null and I don't see any query getting executed in the Toplink to retrieve the streetName.
I don't want to load the complete Address object just to grab the streetName from the Address table.
Did anybody try to do something similar? if yes I would appreciate your hint.
-Rama

TopLink for POJOs does not support lazy loading of primitive type attributes. Lazy loading is only possible with relationships (i.e., one-to-one, one-to-many, many-to-many).
Mapping the same column data to two different classes does entail a number of complications but regardless, you'll eventually have to perform a join between the EMP and ADDRESS table to obtain the street name associated with an Employee. If you model your Employee-Address relationship as a one-to-one rmapping with Employee having an attribute of type Address:
class Emp {
    long id;
    Address address;
class Address {
   long id;
   String streetName;
}you can configure the relationship to use joining[1] so that when you read an Emp you'll also read the associated Address in a single statement.
--Shaun
[1] http://www.oracle.com/technology/products/ias/toplink/doc/10131/main/_html/oomapcfg002.htm#sthref3924

Reg: Loading historic data for the enhanced field

Hello All,
We need to add a new field 0VENDOR to our datasource 0FI_GL_4. This field is available in our BSEG table. Hence, we are planning to go ahead with datasource enhancement.
Now, please advice on how to update the historical data to this newly added field.I have heard there is a BW functionality/program to do so without deleting the entire data. Kindly advice on the possible solutions.
Thanks & Regards
Sneha Santhanakrishnan

HI Sneha,
Using remodeling option you will be able to do that, ie.. loading historical data for new attributes without deleting existing data. But the problem is in remodeling either you can assign constant or some other attribute value or values determined using EXIT.
Now when you are loading data from source system and if you need historical data as well the best practise is delete existing data and reload it from source sytsem.
But if you don't want to do that then I can give you one trick but not sure whether it will work or not. The idea is to populate the historical values for 0Vendor using customer exit option of remodeling. Now to get the historical values in customer exit you will need all that data in some BW table, here you can think of creating some generic extractor which will store the values of all the documents and the respective vendor, as you will load data form source system you will get historical values as well.
Now read that table in customer exit and populate vendor value. This will be one time process to populate the historical values.
Regards,
Durgesh.

One-to-one-to-one direct to field mapping

Yes, I didn't mispelled, there IS a one-to-one-to-one relationship in the database I am mapping, don't ask me why!
To make it simple:
TABLE A (a,b, att1)
TABLE B (b,c, att2)
TABLE C (c, att3)
My Descriptor class is DescA.java with all attributes (String att1, String att2 and String att3) and is associated directly with table A.
I have mapped att1 using direct to field.
I have mapped att2 using direct to field and multitable.
But there is no direct relationship between A and C tables, which makes multitable configuration impossible (?).
So, how can I map C.att3 direct to DescA.att3 ??
[]'s
Henrique

Henrique,
If you are using the Toplink Workbench you will have to specify the relationship between Table B and C using an After Load method. The mapping workbench does not support this scenario yet, but it is allowed by TopLink. Use the api ClassDescriptor.addMultipleTableForeignKeyFieldName(String, String) to specify the relationship.
In the workbench you will see a problem message about the primary keys not matching and no reference being specified. You can safely ignore this warning, or if it bothers you, the only workaround I can think of is to also specify the mappings to Table C in the afterload method.
Karen

SQL-Loader Control file for Tab-delimited fields

i want to import a text-file with TAB-DELIMITED fields and a line-break + return at the end of each record.
a want to do this by sql-loader and a control file.
please, can someone give me the statement:
e.g.
load data
infile 'exaple.txt' .......
(...field1, field2,....)
thank you very much

Case Study 3: Loading a Delimited, Free-Format File
http://otn.oracle.com/pls/db92/db92.to_URL?urlname=http:%2F%2Fdownload-west.oracle.com%2Fdocs%2Fcd%2FB10501_01%2Fserver.920%2Fa96652%2Fch10.htm%231006907
Modify the example and use 'terminated by whitespace' instead
For more information on TERMINATED Fields see
http://otn.oracle.com/pls/db92/db92.to_URL?remark=drilldown&urlname=http:%2F%2Fdownload-west.oracle.com%2Fdocs%2Fcd%2FB10501_01%2Fserver.920%2Fa96652%2Fch06.htm%231013838

6 IF-THEN and OR conditions for one target field mapping

Hi Mapping experts,
I have a validation condition involving 2 source fields and 1 target field only, however the number of conditions is 6 (OR)
Source Fields:
TYPE
PONUM
Target Field:
BELNR
If TYPE = 1st value   then   BELNR = PONUM substring (0, 8) 0=start position, 8=character count
OR
If TYPE = 2nd value   then   BELNR = PONUM substring (2, 8)
OR
if TYPE = 3rd value then BELNR = PONUM substring (4, 8) .......... up to the 6th Validation Condition.
When I do this using graphical mapping only (multiple IF-THEN and OR's) the validation is successful only on the 1st IF-THEN statement, then it passes the substringed value to the 1st OR function.....    When I "Display Queue" the 1st OR function the substringed value is taken as null (or:in1 <null>) Is this because OR function is expecting Input as "True or False" value only and not a string value?
For my requirement above, can I achieve this using graphical mapping only?    or do I have to use UDF for this?
If it's better to use UDF for this, would anyone give me a jumpstart on the code? (I have no experience yet in UDFs)
Thanks, and hoping that somebody will respond accurately and completely.

Create UDF with two inputs as TYPE & PONUM and use bleow code there..
String out = "";
if( TYPE.equals("1"))
out = PONUM.substring(0,8);
return out;
else if( TYPE.equals("2"))
out = PONUM.substring (2,8);
return out;
else if( TYPE.equals("3"))
out = PONUM.substring (4,8);
return out;
else if( TYPE.equals("4"))
out = PONUM.substring (4,8);
return out;
else if( TYPE.equals("5"))
out = PONUM.substring (4,8);
return out;
else if( TYPE.equals("6"))
out = PONUM.substring (4,8);
return out;
return "";
Note: change the substring range as per your need in above code.
Now do your mapping as shown below
TYPE ------>|
            | -----> UDF ---> BELNR
PONUM----->|

About lazy loading

Is there any material about lazy-loading in kodo. If one class contains a
collection type field, how does kodo deal with this field when query the
object ?
Thanks
Liang Zhilong
EMail : [email protected]
Tel : (021)54235858-6650
Fax : (021)54235800
PhotonicBridges Co., Ltd.
12F, 900 Yi Shan Rd., Shanghai 200233

Thanks for your kind help, my Chinese friend.
"Sun Bin" <[email protected]> ??????:[email protected]...
I'm also a JDO user. Glad to see your Chinese name!
I think you should not close the pm before any pending use of PC objects.
You may use a ThreadLocal object to get the pm in different methods
without passing the pm as an argument.
see:
http://www.jdocentral.com/forums/index.php?act=ST&f=11&t=625&s=94c11c31c3acd
a1bf154a6d50fde9b36
>
My answers to your former questions:
Lazy loading is configurable according to JDO spec, to the field-level,
and this is called Default-Fetch-Group. By default, all Collections are
lazy-loaded.
As for the "makeTransientAll" behavior, there's some discussion on
JDOCentral.com/forums/
I remember someone pointed out a solution for your need. While, you can
still use a ThreadLocal object to solve this problem.
Liang Zhilong wrote:
I write a class
public class NciEms
protected String emsVersion = null;
protected String type = null;
protected List subnetworkList = new ArrayList();
protected List managedElementList = new ArrayList();
protected List topoLinkList = new ArrayList();
protected List trafficDescriptorList = new ArrayList();
And an application:
// Obtain pm
pm = xx.getPersistenceManager();
// Query out the jdo object
ems = (NciEms)getJdoObject( pm, NciEms.class,
"name=="/EMS=PhotonicBridges@/SNM"" );
// Print the object
Whether "com.solarmetric.kodo.DefaultFetchThreshold" is set to "30" or
"-1"(Disable lazy loading), the output is always:
1.1) name =/EMS=PhotonicBridges@/SNM
1.2) native ems name=PhotonicVision SNM
1.3) owner =None
1.4) user label =PhotonicVision SNM
1.5) ems version = 1.2.0
1.6) type = MSTP
1.7) me list (length:0) = null
1.8) TL list (length:0) = null
1.9) Subnetwork list (length:0) = null
1.10) TD list (length:0) = null
Why ?
If I close the pm after query the object out, and pass this object to
another method, I cannot use this object in that method because pm has
been
closed. I try to makeTransient the object before it is passed, it doeswork.
But when I use the list field in this object(this list field has been
retrieved out in pm), the "PersistenceManager has been closed" exception
turns up again. Need I makeTransientAll every list fields ? It's awful.
If this usage is required, what the best way?
"Liang Zhilong" <[email protected]> __'____________________
:[email protected]...
Furthermore, what granularity is configurable for lazy loading,
global,
class-level or field-level ?
"Liang Zhilong" <[email protected]> __'____________________
:[email protected]...
Is there any material about lazy-loading in kodo. If one class
contains
a
collection type field, how does kodo deal with this field when query
the
object ?
Thanks
Liang Zhilong
EMail : [email protected]
Tel : (021)54235858-6650
Fax : (021)54235800
PhotonicBridges Co., Ltd.
12F, 900 Yi Shan Rd., Shanghai 200233

Field Mapping of CREMAS05 IDoc..!!

Hi Gurs ,
Iam trying to map the following fields in CREMAS05 Basic IDoc Type but unable to find the correct mapping field segments in the IDoc.
Fields need to be mapped are:
1.Building Code
2. Pan No
Can anyone suggest me which segment and field i need to use for the above fields mapping..? or if i extended my idoc with some custom table how i will map into Standard field .
Thanks .
Hemant Gulati(A1)

Hi Jurgen ,
Thanks for Reply .
Vendor Master Idoc CREMAS05
i have extended my idoc with some zfields . in zfield i have Building code and Pan no
like zbuild and zpanno i need to map these in ADDR1_DATA-BUILDING and J_1IMOVEND-J_1IPANNO respectively .
Thanks & Regards
Hemant Gulati(A1)

Catalog integration in MM (html - SAP-field mapping)

Hi Experts! When integrating Catalog with MM (ME21N and ME51N), the only customizable entry in the IMG is the call-structure. Anyone who knows the transaction codes / tables for HTML-SAP-field mapping (OCI) for this scenario ?? (For PM/CS/PS-orders this is available, but I have not managed to find this for MM)...
brgs ziggy

to pull Metadata from CMS repository into MM, you will have to configure the CMS Integrator and run that CMS Integrator, this you can do from the CMC
as Werner mentioned SAP BW Integrator is being added for next MM release
its not possible to import Metadata from Excel Sheet into MM, you will need a Integrator to do that, MM supports Integrator for CMS, Data Services, RDBMS, and there are few more Integrators that are provided by MITI (it requires a separate license)
I don't think it will be possible to mimic the impact and lineage

JPA - lazy loading for LOB field

Hi all,
JPA 1.0 specification mandates that all JPA-compliant implementation support lazy loading for certain kind of entity field.
For LOB fields lazy loading is OPTIONAL.
I am experiencing odd runtime behaviors on my custom software which would point to this feature not being supported.
Can anyone please tell me if SAP JPA 1.0 implementation on NW CE 711 implements this feature or not?
Thanks in advance
Regards
Vincenzo

Hi Vincenco,
I am sure that this is the same as with single-valued relationships (@OneToOne, @ManyToOne): Lazy loading would require bytecode manipulation/generation, so SAP JPA does not support it in 7.20 (and of course not in 7.11)
See tulsi jiddimani's elaborate answer here: Re: JPA: Documentation on LazyLoad.
In 7.30 enhancements, you really can find lazy loading support for single-valued relationships with getReference.
http://help.sap.com/saphelp_nw73/helpdata/en/68/f676ef36094f4381467a308a98fd2a/content.htm
but @Lob and @Basic is not mentioned.
If you need lazy loading in 7.11, you have two alternatives:
1. Put the Lob fields into separate entities, work around the missing feature in SAP JPA with ugly @OneToMany - Relations
2. Use another persistence provider like EclipseLink, read Sabine Heider's blogs about integrattion of EclipseLink in SAP NetWeaver and static bytecode weaving for lazy loading. /people/sabine.heider/blog
Regards
Rolf

Is there a way for a form user to load graphics and text in same field?

I am aware that you can load text into a text field and graphics into an image field. However, I would like to load a document containing text and graphics (e.g., MS Word or RTF) into a single field on a PDF form. Is there any way to do this?

Paul,
I believe that the example of attaching a file that you have provided will solve the problem. I am using LiveCycle Designer, so using the code that is in your example should be no problem. The end-user will use Adobe Reader, so as you mentioned, I will use Acrobat to extend the features so that the form can be used by Reader.
Thanks so much for your help,
Daniel
From: pguerett <[email protected]>
To: Daniel Beinart <[email protected]>
Sent: Monday, April 25, 2011 3:03 PM
Subject: Re: Is there a way for a form user to load graphics and text in same field?
It is part of the PDF AcroForm side of things .....Are you using Reader or Acrobat to fill the form. If Reader then you will have to Reader Extend to make this work ....if Acrobat then you simply need to add some code. I have attached a sample that shows you how to do it if you want it.
Paul

Toplink.weaving=true gives PersistenceException; for Lazy Loading config.

I could use some help configuring my persistence.xml file, trying to make sure Lazy loading is configured properly. What I see in the logs normally is that Toplink defaults to Eager fetching on my JPA objects. So according to
http://www.oracle.com/technology/products/ias/toplink/JPA/essentials/toplink-jpa-extensions.html
I need to set toplink.weaving=true in persistence.xml, however I get the Exception below when I make the call to get the EntityManager for the first time.
I'm using OC4J standalone 10.1.3.3 with JDK 1.5.0_13 and the latest Toplink (as you can see in the stack trace below). An excerpt from my persistence.xml is below that.
Caused by: Exception [TOPLINK-30005] (Oracle TopLink Essentials - 2.0 (Build b58g-fcs (09/07/2007))): oracle.toplink.essentials.exceptions.PersistenceUnitLoadingException
Exception Description: An exception was thrown while searching for persistence archives with ClassLoader: xyx-dal-webservice.root:0.0.0
Internal Exception: javax.persistence.PersistenceException: Exception [TOPLINK-28020] (Oracle TopLink Essentials - 2.0 (Build b58g-fcs (09/07/2007))): oracle.toplink.essentials.exc
eptions.EntityManagerSetupException
Exception Description: Value [true] for the property [toplink.weaving] is incorrect when global instrumentation is null, value should either be null or false. at oracle.toplink.essentials.exceptions.PersistenceUnitLoadingException.exceptionSearchingForPersistenceResources(PersistenceUnitLoadingException.java:143)
at oracle.toplink.essentials.ejb.cmp3.EntityManagerFactoryProvider.createEntityManagerFactory(EntityManagerFactoryProvider.java:169)
at javax.persistence.Persistence.createEntityManagerFactory(Persistence.java:59)
at javax.persistence.Persistence.createEntityManagerFactory(Persistence.java:48)
at gov.dhs.visit.ident.dal.jpa.entity.cvt.util.EntityManagerHelper.<clinit>(EntityManagerHelper.java:20)
... 43 more
Caused by: javax.persistence.PersistenceException: Exception [TOPLINK-28020] (Oracle TopLink Essentials - 2.0 (Build b58g-fcs (09/07/2007))): oracle.toplink.essentials.exceptions.E
ntityManagerSetupException
Exception Description: Value [true] for the property [toplink.weaving] is incorrect when global instrumentation is null, value should either be null or false.
at oracle.toplink.essentials.internal.ejb.cmp3.JavaSECMPInitializer.callPredeploy(JavaSECMPInitializer.java:151)
at oracle.toplink.essentials.internal.ejb.cmp3.JavaSECMPInitializer.initPersistenceUnits(JavaSECMPInitializer.java:239)
at oracle.toplink.essentials.internal.ejb.cmp3.JavaSECMPInitializer.initialize(JavaSECMPInitializer.java:255)
at oracle.toplink.essentials.ejb.cmp3.EntityManagerFactoryProvider.createEntityManagerFactory(EntityManagerFactoryProvider.java:155)
... 46 more
Caused by: Exception [TOPLINK-28020] (Oracle TopLink Essentials - 2.0 (Build b58g-fcs (09/07/2007))): oracle.toplink.essentials.exceptions.EntityManagerSetupException
Exception Description: Value [true] for the property [toplink.weaving] is incorrect when global instrumentation is null, value should either be null or false.
at oracle.toplink.essentials.exceptions.EntityManagerSetupException.wrongWeavingPropertyValue(EntityManagerSetupException.java:244)
... 50 more
SEVERE: Caught exception while handling request: oracle.oc4j.rmi.OracleRemoteException: java.lang.ExceptionInInitializerError; nested exception is:
java.lang.ExceptionInInitializerError oracle.oc4j.rmi.OracleRemoteException: java.lang.ExceptionInInitializerError; nested exception is:
java.lang.ExceptionInInitializerError
persistence.xml
<persistence-unit name="PUName" transaction-type="JTA">
     <provider>oracle.toplink.essentials.PersistenceProvider</provider>
          <jta-data-source>jdbc/mydb</jta-data-source>
<class>....a whole lotta JPA classes...</class>
    <properties>
         <property name="toplink.logging.level" value="FINE" />
         <property name="toplink.logging.timestamp" value="true"/>
           <property name="toplink.target-database" value="Oracle"/>
           <property name="toplink.target-server" value="OC4J_10_1_3" />
           <property name="toplink.weaving" value="true"/>
                <property name="toplink.cache.shared.default" value="false"/>
    </properties>
</persistence-unit>

This error confuses me...
Exception [TOPLINK-30005] (Oracle TopLink Essentials - 2.0 (Build b58g-fcs (09/07/2007))): oracle.toplink.essentials.exceptions.PersistenceUnitLoadingException
Exception Description: An exception was thrown while searching for persistence archives with ClassLoader: sun.misc.Launcher$AppClassLoader@11b86e7
Internal Exception: javax.persistence.PersistenceException: Exception [TOPLINK-28020] (Oracle TopLink Essentials - 2.0 (Build b58g-fcs (09/07/2007))): oracle.toplink.essentials.exceptions.EntityManagerSetupException
Exception Description: Value [true] for the property [toplink.weaving] is incorrect when global instrumentation is null, value should either be null or false.
     at oracle.toplink.essentials.exceptions.PersistenceUnitLoadingException.exceptionSearchingForPersistenceResources(PersistenceUnitLoadingException.java:143)
     at oracle.toplink.essentials.ejb.cmp3.EntityManagerFactoryProvider.createEntityManagerFactory(EntityManagerFactoryProvider.java:169)
     at javax.persistence.Persistence.createEntityManagerFactory(Persistence.java:110)I'm using TopLink from regular ol' Java SE (not container managed environment).
http://www.oracle.com/technology/products/ias/toplink/JPA/essentials/toplink-jpa-extensions.html gives some hints, though, especially if you're operating outside of a container (seems like you gotta weave yourself). See http://www.oracle.com/technology/products/ias/toplink/JPA/essentials/toplink-jpa-extensions.html#LazyLoading
Seems the command-line arg "-javaagent:toplink-essentials-agent.jar" is critical if you're running outside a container.

Historical data load for newly added fields

Hi,
we are using cube 0UCSA_C01 with delta update which has gone live now the requirement is we have to add 2 new fields to this cube but the problem is how can we load previous data for dese fields dat has been updated already . Plz letme know hw this can b acheieved plz letme know the proceedure to achieve dis..
Thnx
help will b apprciated

Hi,
populate historical data into a newly added field in an infoprovider
How to populate historical data into a newly added field in an infoprovider
/people/dinesh.lalchand/blog/2006/02/07/loopback-process--follow-up
http://sap.ittoolbox.com/documents/popular-q-and-a/adding-characteristics-to-infocubes-2138
Thanks,
JituK

Disable Lazy Loading for Navigation Properties on Entities in EF 5

Hello all,
Can we disable Lazy Loading for Navigation properties on any entity?
If yes, how we can achieve the same.
Thanks,
Naresh

Hi Naresh;
If you are using code first you can do the following to turn off Lazy loading for a given navigation property. For example
lets say we have a Customers table and a Orders table and a Customer can have 1 to many Orders the following POCO class has a ICollection that is mark as virtual, This causes Lazy loading to happen when a property of order is accessed.
// Lazy Loading Enabled
public class Customers
public int Id { get; set; }
public string Name { get; set; }
public virtual ICollection<Order> Orders { get; set; }
To turn off Lazy loading remove the virtual key word as follows:
// Lazy Loading Is Disabled
public class Customers
public int Id { get; set; }
public string Name { get; set; }
public ICollection<Order> Orders { get; set; }
To turn of Lazy Loading for all entities using Code First you can do the following in the constructor of the DbContext.
public class MyContext : DbContext
public MyContext()
this.Configuration.LazyLoadingEnabled = false;
If the above does not meet your needs please tell us what version of the Entity Framework, which context type you are using, for example ObjectContext or DbContext, and which modeling you are using such as Code First, Model First or Database First.
Fernando (MCSD)
If a post answers your question, please click "Mark As Answer" on that post and "Mark as Helpful".

Load the data for which these 2 fields are not initial (Blank).

Hi Experts,
I am extracting data from R/3 to BW using a Z data source based on view.
In the extract there will be 2 fields User ID & Date.
My requirement is to load the data for which these 2 fields are not initial (Blank).
My Data source is extracting all the data from R/3 to BW.How can filter it out in BW.
Thanks

Hi Siegfried,
Thanks for the info Provided.
As I am new to BW, Could you please provide me with the code (ABAP) I need to insert and where...in which part of the below start routine in need to insert the code....
My requirement is to load the data for which the USER ID (ZUSR_ID) & DATE (ZDATE) is not null/blank/0.
<b>Start Routine IN Transfer Rules</b>
Global code used by conversion rules
$$ begin of global - insert your declaration only below this line -
TABLES: ...
DATA:   ...
$$ end of global - insert your declaration only before this line   -
FORM STARTROUTINE
USING    G_S_MINFO TYPE RSSM_S_MINFO
CHANGING DATAPAK type TAB_TRANSTRU
           G_T_ERRORLOG TYPE rssm_t_errorlog_int
           ABORT LIKE SY-SUBRC. "set ABORT <> 0 to cancel datapackage
$$ begin of routine - insert your code only below this line        -
DATA: l_s_datapak_line type TRANSFER_STRUCTURE,
      l_s_errorlog TYPE rssm_s_errorlog_int.
abort <> 0 means skip whole data package !!!
ABORT = 0.
$$ end of routine - insert your code only before this line         -
ENDFORM.
Thanks for the help.

Lazy loading or indirection for direct to field mapping striing object

Similar Messages

Maybe you are looking for