Partial Update - Unknown Dimension Value ID

Hello Old Endeca Friends,
I am getting a lot of these sorts of errors when I perform a partial update:
ERROR 04/25/13 16:56:21.115 UTC (1366908981115) DGRAPH {dgraph,update} Record change references unknown dimension value id 4294484500 [recspec=`UPI:100224148']
I am using MDEX 6.4 and PS 6.1.3. The dimension value ids in question seem to be in the autogen_dimensions.xml, but the dgraph doesn't seem to recognize to know about them. Why?
Any further insight would be appreciated. "Just run a baseline" is not a good answer, I need to do partials every hour throughout the day.

Thanks for the input, everyone.
It looks like my issue was that my state file had a set of updated dimensions but my dgraph did not.
I had a baseline extract with properties X, Y and Z (let's say) and I was adding in new dimensions A and B. I moved my state files from my test environment to my production environment with entries on dimensions A and B. I started getting partial data for dimensions A and B and processing it. However, my dgraph knew nothing about A and B. So, although forge created the dim vals fine and the state files were modified appropriately, my dim val updates and adds on my dgraph failed because my dgraph had never processed a baseline extract with dimensions A and B.
Once I ran a baseline with dimensions A and B, partial updates went fine and have now processed dozens without issue.
Unfortunately, I only get baseline data once a week, so it took me a bit of time to realize what was going on.

Similar Messages

Error in Partials for new dimension values

Hello everyone,
I was trying out partial update for the first time on our index. ADDS, UPDATES, DELETES everything works out fine. The only error we are seeing is when an ADD or UPDATE has a new dimension value in it, the following error shows up:
*DGRAPH {dgraph,update} Record change references unknown dimension value id 4294428511 [recspec=`P_RECORD_SPEC:GS1001B101508E']*
This is what I observe:
~ data/state/autogen_dimensions.xml has the entries for this dimension after partial forge is completed
~ the forge output files SomeName--sgmt0.records.xml has the entries for this dimension
My thinking is the new dimensions collected in autogen-dimensions.xml during the partial are not available to the dgraph when the update is run. Am I missing any steps? Can you please point me the right direction. Please let me know if you need additional information on my setup.
[Product: Endeca Information Access Platform version 6.1.4.530518 (analytics enabled)]
Regards.
Pankaj.

Can you share your XML record manipulator where you tag each record with the update action? This would be a place where you would see:
<EXPRNODE NAME="ACTION" VALUE="ADD_OR_REPLACE"/>
The missing piece might be: <EXPRNODE NAME="DIM_ACTION" VALUE="ADD"/>
The latest documentation has a good, dedicated partial update guide.
[MDEX 6.3.0 Partial Update Guide|http://docs.oracle.com/cd/E35825_01/MDEX.630/pdf/PartialUpdatesGuide.pdf]

Dimension values display order and record details

Hi All,
I have explicitly ordered my dimension values to be in a logical order instead of alphabetical. Ex: Small, Medium, Large. This works great in the navigation tree but when I view a specific record and it has those 3 dimvals they are listed in alphabetical order.
Is there a when doing a record query to ask Endeca to respect the same display order as in the nav query?
Thanks,
Luc

We are used to add two dummy records in each dimension.
We add the following data:
We use '-1' (surr-key) for Invalids and '-2' for Unknown.
Easy way to populate these dimensions is to add a script together with all objects when you release new functionality. Beware you use the same descriptions.
Even better is to think of a more general solution (more easy to maintain) and store these values in a support table. In such a table you will be able to add more columns to get more details per row. Use this reference table when you load youre dimension. Also you will find the opportunity to add more error handling cases. Especially when you add the dimension in this support table.
Regards
Moscowic

Is partial update possible?

Hi,
is it possible to have Toplink perform partial update of an object, instead of all the attributes. What I need is exactly the same feature as Toplink's support for partial reading where you can specify which fields to retrieve from database.
It's necessary to disable the update of some fields in some circumstances even if the field's value is different from what is in database row. it sounds meaningless but needed.
regards & thanks.

I am confused by your request. TopLink's UnitOfWork provides minimal update capabilities. It will only modify those attributes that are different from when the object was read in.
Doug

Update dynro field values in PAI of BTE event 00000920

Hello
I have the following specific requirement in transaction f-28 : when I click a specific button called 'COMPUTE', I need tu update payment amount in tab 'Partial payment' with a specific amount instead of the default amount.
So I have created the button 'COMPUTE' using BTE event 00000930 and it's OK.
the specific amount is computed using BTE event 00000920 in a specific FM. In this FM, after the amount is known, i want to display it on the screen.
The field payment amount in dynpro is DF05B-PSZAH (dynpro nb for batch-input is 3100, the dynpro number of the program is 6104).
How can I display the computed value in this field of dynpro ?
the last thing is : when I change manually (without button 'COMPUTE') the payment amount, the field 'assigned' is automatically re-compute (PAI screen 6104).
How can I trigger this automatic change after having updated the dynpro field 'DF05B-PSZAH' ? In other way, how can i trigger this automatic change ?
Thanks all

I've found myself the technical solution : I used a enhancement point in FORM pos_zeile_bearbeiten of the program SAPDFF05X.
First I still use BTE event 00000930 to create the specific button.
To update the partial payment :
- set to 'X' the field xaktp in the table postab for the line of the item
- set to 'X' the field inpud in the table indtab for the line of the item
- change the value of the field pszah (amount of the partial payement) and the value of the field psdif (diffrence between the amount of the invoice and the amount of the payment)
Caution : the number of lines is not the same in postab and indtab (postab contains all items even those not shown in the screen and indtab contains all items shown in the screen) - to find the corresponding line inf ind tab ; search in table tsltab the value of the posnr corresponding to the item and the sy-index of this search will give you the index inn table indtab for the same item.

Execution of a planning function with the used dimension values in a crosstab

Hello,
would like to execute planning functions within Design Studio but I don't really know how I combine this execution with the used dimension values in a crosstab for the .
Here the requirement:
3 Planning functions which should use the same filter (all values of all dimensions in the crosstab) and which should run one after another:
// open data slice
PF_1.execute();
// generate combinitions
PF_2.execute();
// close data slice
PF_3.execute();
Many thanks in advance.
Kind regards,
Tobias

Hi Arun,
in general it is not possible to nest planning functions. The inner function would update the data in the plan buffer without the outer function knowing about it. Also the inner function might be on a different aggregation level which would make things very tricky.
Alternatively you can include ABAP logic into a FOX formula:
http://help.sap.com/saphelp_nw70/helpdata/en/47/94eb78b6394befe10000000a42189d/frameset.htm
Or just create the complete logic in one ABAP planning function type.
Regards,
Marc
SAP NetWeaver RIG

Partial Update error / sample code

Customer is trying to do Partial Update by refering the sample shown in PartialUpdatesGuide.pdf, however, baseline update is showing following error. Is there any sample code to do partial update with the actual success in some instance, or any resolution for this ?
WARNING:provisioning warning: The option '--updatedir' is reserved. It has been removed from the command line
Additional informations are:
- after partial update, dgraph_input/updates does not have partial update file (directory is empty). It seems that they does not use Parial Update Pipeline in forge, this file is not created.
- Dgidx log does not show any other error or warning

The message you see is just a warning that indicates the Dgraph node in your AppConfig XML (either DgraphCluster.xml or AuthoringDgraphCluster.xml) is not specified properly. My guess is that yours looks like this:
<dgraph id="DgraphA1" host-id="LiveMDEXHostA" port="25000"
post-startup-script="LiveDgraphPostStartup">
<properties>
<property name="restartGroup" value="1" />
<property name="DgraphContentGroup" value="Live" />
</properties>
*     <args>*
*          <arg>--updatedir</arg>*
*          <arg>./data/dgraphs/DgraphA1/dgraph_input/updates</arg>*
*     </args>*
<log-dir>./logs/dgraphs/DgraphA1</log-dir>
<input-dir>./data/dgraphs/DgraphA1/dgraph_input</input-dir>
+<update-dir>./data/dgraphs/DgraphA1/dgraph_input/updates</update-dir>+
</dgraph>
The error message is complaining about the args section. The updatedir argument gets sourced from the update-dir element and if you try to specify it in args, you see the message below.
I think we'll need more information to solve the primary issue (what does your PartialForge.log say, is there anything in data\partials\forge_output). Dgidx does not run during a partial update so not having a warning or error there is expected.
Regards,
Patrick Rafferty
http://branchbird.com

Unknown Dimension Name in Keyword "0):    LOOKUP function

Hi,
I am doing Migration project. from BPC MS to NW 7.5 In this I have vallidated the logics in the NW version as per MS version.
I got only one error when I validate the LOOKUP function in the NW under the same application(Staff_Additions).
The error message: "Unknown Dimension Name in Keyword "0):"
Could you please help me in this and please find the code below.
*Lookup Staff_Additions
*DIM SALARY="Salary_redis"
*DIM CATEGORY="Budget"
*DIM CLIENT_GROUP="1010"
*DIM DATASRC="Input"
*DIM INTCO="Non_Interco"
*DIM RPTCURRENCY="LC"
*DIM TIME="2011.MAR"
*DIM ACCOUNT="41000"
*ENDLOOKUP
*WHEN CATEGORY
*IS = "BUDGET"
*WHEN INTCO
*IS = "Non_InterCo"
*WHEN JOB_TITLE.ID
*IS = "No_JobTitle"
*WHEN SALARY
*IS = "Salary_Redis"
*WHEN RPTCURRENCY
*IS = "LC"
*WHEN DATASRC
*IS = "INPUT"
REC(EXPRESSION=ROUND(.07692 Lookup(Staff_Additions),0),Time="2010.APR",SALARY="SALARY_MO")
*ENDWHEN
*ENDWHEN
*ENDWHEN
*ENDWHEN
*ENDWHEN
*ENDWHEN
*COMMIT
Thanks and Regards,
Krishna.

Hi,
My intension is multiple with .06792 and make it round figure and store the value in 2010.MAR.
I have modified this logic in the NW Version,It shown below.
*Lookup Staff_Additions
*DIM SALARY="Salary_redis"
*DIM CATEGORY="Budget"
*DIM CLIENT_GROUP="1010"
*DIM DATASRC="Input"
*DIM INTCO="Non_Interco"
*DIM RPTCURRENCY="LC"
*DIM TIME="2011.MAR"
*DIM ACCOUNT="41000"
*ENDLOOKUP
*WHEN CATEGORY
*IS = "BUDGET"
*WHEN INTCO
*IS = "Non_InterCo"
*WHEN JOB_TITLE
*IS = "No_JobTitle"
*WHEN SALARY
*IS = "Salary_Redis"
*WHEN RPTCURRENCY
*IS = "LC"
*WHEN DATASRC
*IS = "INPUT"
REC(EXPRESSION=.07692(Staff_Additions),TIME=2010.APR,SALARY=Salary_MO)
*ENDWHEN
*ENDWHEN
*ENDWHEN
*ENDWHEN
*ENDWHEN
*ENDWHEN
*COMMIT
Let me know is this the right procedure to get the right result.
Could you please help me in this.
Thanks and Regards,
Krishna.

Record Store Merger with partial Update

Hi,
I have a requirement in which I have to merge a feed into the endeca via partial update. The feed will only have 3 properties- part number , vendor name , Store id(value to be merged). The combination of part number and vendor name is not unique.
The only to achieve the sdesired result is by merging the feed's store detail in the data base. And by incremental indexing pushing the data into endeca. Somebody told me that CAS will then provide me the difference data and with that difference data I can partially update the records.
But my client doesn't want to use Simple indexing. She is asking to figure out merging data in CAS itself. I know that we can use Record Store Merger to merge two or more data source.
But if I use record Store merger will it provide me same differencial data input as provided by CAS in the case of indexing??
Also I am using Endeca version 6.4.0.

apparently these is a known bug in Endeca which occurs when an update results in a product not falling in 1 of the dimensionvalues. So if the price changes from 100 to 1mln and the price range only goes to 1000, endeca does not update correctly

DESTINATION_APP dimension value unable to be set in *REC

Hi,
I want to move data from one application to another and set the value of a destination dimension (which is not part of the source application so I use ADD_DIM to add it) to the property of a source dimension.
However, I get a validation error at the *REC statement stating the destination dimension is not a valid keyword. All dimensions which are NOT part of ADD_DIM work in *REC but all new dimensions which are listed in ADD_DIM do not work in the *REC statement.
Here is the code I am trying to validate:
*DESTINATION_APP = GL
*SKIP_DIM = Z_COSTCENTRE, Z_CX_IDNO, Z_CX_KEYFIG, Z_FLOW, Z_LCGU
*ADD_DIM Z_GLACCOUNT=31301, Z_GL_ENTRYNO = CX001, Z_GL_KEYFIG = DR, Z_GL_BU = BU_1070
*WHEN Z_CX_KEYFIG
*IS "CX_SPEND"
*REC(FACTOR=1,Z_GL_BU=Z_COSTCENTRE.BU)
*ENDWHEN
The error says: "Unknown dimension name in Keyword: Z_GL_BU"
Any idea's on how I can set the value of the destination based on a property of the source.

Hi,
May be you can try:
*DESTINATION_APP = GL
*SKIP_DIM = Z_COSTCENTRE, Z_CX_IDNO, Z_CX_KEYFIG, Z_FLOW, Z_LCGU
*ADD_DIM Z_GLACCOUNT=31301, Z_GL_ENTRYNO = CX001, Z_GL_KEYFIG = DR, Z_GL_BU = BU_1070
*WHEN Z_CX_KEYFIG
*IS "CX_SPEND"
*REC(FACTOR=1)
*ENDWHEN
The reason I removed the other part from the REC statement is because you are adding the dimension Z_GL_BU with a default value of BU_1070. This means that every record transferred will have the same BU_1070. We cannot change the member of this dimension using the REC statement.
Hope this helps.

How to update the table value in the valuechange event?

I have an input field in the datatable with the valueChangeListener
<rich:dataTable id="cart" value="#{cart.cartList}" var="item">
<h:inputText value="#{item.cost}" id="qty" valueChangeListener="#{items.updateCost}" onchange="submit()">
<h:outputText value="#{item.errorMsg}"> </h:outputText>
in the backing bean
     Item item = (Item) model.getRowData();
// do some update, if the cost too larger, change to max_cost
     item.setCost(max_cost);
     item.setErrorMsg("Error Msg");
After calling the valuechange method, the screen output doesn't update the cost.
How to update the table value in the valuechange event?

As you're misusing the valueChangeListener to set another input field, you need to skip the update model values phase. Otherwise the value set in the valueChangeListener will be overridden by the submitted value. You can do this by calling the FacesContext#renderResponse() inside the valueChangeListener method. This will shift the current phase immediately to the render response phase, hereby skipping the update model values and invoke application phases.

How to update excise invoice values after migo.

Hi Guru's,
I have checked the forum. but did not get exact solution.
Here im having the problem with bapi BAPI_GOODSMVT_CREATE . This bapi don't have the structure to update excise invoice values.
I have one bapi 'BAPI_EXCINV_CREATE_FROMDATA' to update the excise duty, but in that also i didn't find the structure
'J_1IEXITEM' to update the ECS value ( field is 'total_ecs').
Please help me on this,Thanks in Advance...
Regards
Naren

Hi Krishna
Do not do excise invoice capture during 103. At that time select the option "No excise entry". Then capture excise invoice during 105.
The system messge " No selectable items for this document" is prompted since the preceeding document is already assigned with invoice.
Try this and give feedback.
regards
sairam akundi

How to limit dimensions values of querybuilder

Hi,
I want to learn how to limit dimension value in query builder. i want, the user who run the querybuilder can not see a dimension value which i do not want to display.
For example,
The time dimension has "2000" and "2001" values as default which is defined in database. Normally when i run the querybuilder for create new presentation or calculation, In the dimensions panel of the querybuilder display all values of each of dimension. The user can select values which she/he want from that list.
But i want to limit values which the user can see in the dimensions panel of the querybuilder. for this time dimension example, i want to hide "2001" value from the user. How can i do it. if there is any sample code, please share.
Thank you very much for your helps.
ilknur

Thanks Thomas, excellent suggestions as always.
If anybody wants more information regarding Thomas' suggestions please refer to
Oracle9i OLAP Developer's Guide to the OLAP DML for the AW Permit command.
In sumamry: To create permission programs, you define two programs with the names permit_read and permit_write. In these programs, you can specify PERMIT commands that grant or restrict access to individual workspace objects. In addition, you write these programs as user-defined functions that return a Boolean value, and the return value indicates to Oracle OLAP whether or not the user has the right to attach the workspace.
For relational security there is an excellent tutorial on the database pages of OTN, follow this link:
http://otn.oracle.com/obe/start/index.html
then follow the links for Oracle9i Database Release 9.2.0.2 - Security - Creating Label-based Access Control. This module describes how to use Oracle Label Security to setup security based on label policies
These security layers should be transparent to the OLAP metadata layer, therefore, once you have implemented your chosen security method your BI Beans application will only need to connect with the appropriate user to inherit the security layer. For more information see the Security section of the BI Beans Help:
http://otn.oracle.com/bibeans/903/help/
BI Beans product Management Team
Oracle Corporation

How to update a string value located on the main VI

Hello
I would like, from a subVI, to update a string value located in the main VI (using References). I was trying to find examples on NI web page but it´s impossible to download them (for more than one month : Error, File Not Found), do everyone has got the same problem?
Thanks for your answer
Nicolas

Hi,
this is from LV help file:
Complete the following steps to create a subVI that includes a control reference directly from a Property Node.
1.On the block diagram, right-click the front panel object terminal for which you want a control reference and select Create»Property Node from the shortcut menu to create a Property Node implicitly linked to the object. You can have multiple Property Nodes for multiple front panel objects.
2.Use the Positioning tool to select the new Property Node and select Edit»Create SubVI from menu. LabVIEW automatically creates a control reference to the control on the block diagram of the main VI, creates a subVI that contains a control refnum wired to a Property Node, and wires the control reference on the main VI to the
control refnum input terminal of the new subVI.
Note: Select several Property Nodes at once to create a subVI with multiple control refnums and Property Nodes.
3.Double-click the subVI to open it. Modify the Property Node and other elements of the subVI as you do in any other VI.
4. If you want to change values of control then select "Value" property in Property node.
I've attached the examples.
Good luck.
Oleg Chutko.
Attachments:
Ref.zip ‏11 KB

Error while updating the forecast value in M.Master through RMDATIND

Hi All
Good Day
In my custom Program I am using the three structures bgr00,bmm00,bmmh1 & bmmh3 and through rmdatind(Direct input method) program, i am trying to update the forecast value in material master data
and i am getting error message as Specify a tracking limit and i have taken the tracking limit as 4 in my program.
Actually there are 2 fields exists for tracking limit in structure bmmh1
1.SIGGR and
2.SIGG2
I had given the input as 4 for these 2 fields but I am getting the same error message Specify a tracking limit
Can anyone please help me in this case.
Thanks in advance.
Cheers
Vallabhaneni

debug your program.
Compare the field content of the value in the field you submitted with the content in field that is checked when message M3 456 is issued.

Partial Update - Unknown Dimension Value ID

Similar Messages

Maybe you are looking for