Trex 7.0 index issue

Hi All,
In our portal system which is on SAP Netweaver 7.0 ehp1 with Trex 7.0 version, there is a strange behaviour.
The index corresponding to the repositories although they get created on reindexing every time allow for searching for the documents on the portal for few hours only.
After few hours , the number of documents indexed is greatly reduced at times to even 0 and hence the search does not yield any results.
Hence, the search only works few hours from the point the reindex is done and stops working after few hours.
Any suggestions for resolving this strange behaviour is highly appreciated.
Regards,
Murali Narayanan

Hi Murali,
Can you give us a little more information about the problem that you're seeing? Is this happening to just one index or multiple? What type of repositories are these that you are indexing? Is there a rebuild schedule set up on the index so that they automatically rebuild every day?
Looking at the TREX monitor, how many documents are showing up in the 'OK' column for your index?
-Stephen Spalding

Similar Messages

TREX is not indexing certain excel files

Hi Experts,
TREX is not indexing certain excel-files correctly. I get a "no message found error" (14047) Trex: Preparation failed: index operation by certain excel files - but not by all. I found out what they have in common: some have filters activated and some have a lot of columns or rows.
Is there any way to avoid these errors and index them correctly?
TREX: 7.0.47
Portal: 7.01 SP7
Thank you in advance
Best Regards
Norman

Hi Norman,
Did you review note 1528705? It has a fix for this issue.
Best regards,
Duncan

BIA index issues.

Hi All,
i have an BIA-Index issue,The query-result is wrong when i executes with BIA-Index in transaction RSRT .
But when i checked the master data of infoobject is consistent. I have checked this in transaction RSRV.
Could you please give any ideas how to correct this?
Thanks,
Vikram.

Hi,
Most likely you need to run an attribute change run. Attribute change run updates BIA index according to data stored in a InfoObject. You can do it from both a Process Chain or manually through Administrator Workbench -> Tools -> Hierarchy/Attribute Changes.
Regards,
Adam

Indexing Issue : Idc Analyzer and other tools

Hi All,
I am facing some indexing issue with my ucm instance. Some of the files get stuck in wwGen Revision Status, some show “up to date” in index, but really are not until a re-index
happens, etc.
I used IDCAnalyzer to check the indexing issues but during analysis i got only an error pop-up saying "Error checking index" with no details in log.
Are there any additional arguments/settings which can be used to get the details. Otherwise, this tool doesn't seem to be of much help in this case.
Which are the other tools that can be used to check and correct the health of index?
How can we purge unneeded revisions, history, and other in UCM to “clean up” and remove bloat (like archiver etc.)
Note: I am using ucm 11g with SSXA.
Edited by: PoojaC on Aug 1, 2012 10:26 PM

Hi ,
Analyzer tool should be used when there is a mismatch in the weblayout / vault files which causes indexer , archiver etc to fail .
Read more about this from the following links:
http://docs.oracle.com/cd/E23943_01/doc.1111/e10792/e01_interface.htm#CACFDIID
http://docs.oracle.com/cd/E23943_01/doc.1111/e10792/c03_processes.htm#sthref268
Hope this helps .
Thanks
Srinath

LSMW -Recording Index issue

Hi,
I was trying to add partners to a customer through XD02.
I need to add multiple Partners to a customer.
I recorded transaction but I have index issue .
For example when I recorded it is 6th line where i need to enter but for next record it should add in 7th line.
but 6th line is getting repalced.. how do i resolve this issue?
Should I code something in FIELDMAPPING?
Regards
Prasad

Hai Vara
I am giving Material master Upload through LSMW Direct Input Method
Just Follow The Steps
Using Tcode MM01 -- Maintain the source fields are
1) mara-amtnr char(18)
2) mara-mbrsh char(1)
3) mara-mtart char(4)
4) makt-maktx char(40)
5) mara-meins char(3)
the flate file format is like this as follows
MAT991,C,COUP,Srinivas material01,Kg
MAT992,C,COUP,Srinivas material02,Kg
AMT993,C,COUP,Srinivas material03,Kg
MAT994,C,COUP,Srinivas material04,Kg
MAT995,C,COUP,Srinivas material05,Kg
goto Tcode LSMW
give Project Name
     Subproject Name
     object Name
Press Enter -
Press Execute Button
It gives 13 radio-Button Options
do the following 13 steps as follows
1) select radio-Button 1 and execute
   Maintain Object Attributes
select Standard Batch/Direct Input
   give Object -- 0020
       Method -- 0000
   save & Come Back
2) select radio-Button 2 and execute
   Maintain Source Structures
   select the source structure and got to click on create button
   give source structure name & Description
   save & Come Back
3) select radio-Button 3 and execute
   Maintain Source Fields
   select the source structure and click on create button
   give
   first field
        field name    matnr
        Field Label   material Number
        Field Length 18
        Field Type    C
   Second field
        field name    mbrsh
        Field Label   Industrial Sector
        Field Length 1
        Field Type    C
   Third field
        field name    mtart
        Field Label   material type
        Field Length 4
        Field Type    C
   fourth field
        field name    maktx
        Field Label   material description
        Field Length 40
        Field Type    C
   fifth field
        field name    meins
        Field Label   base unit of measurement
        Field Length 3
        Field Type    C
save & come back
4) select radio-Button 4 and execute
   Maintain Structure Relations
   go to blue lines
      select first blue line and click on create relationship button
      select Second blue line and click on create relationship button
      select Third blue line and click on create relationship button
save & come back
5) select radio-Button 5 and execute
   Maintain Field Mapping and Conversion Rules
   Select the Tcode and click on Rule button there you will select constant
   and press continue button
   give Transaction Code : MM01 and press Enter
   after that
   1) select MATNR field click on Source filed(this is the field mapping) select MATNR and press Enter
   2) select MBRSH field click on Source filed(this is the field mapping) select MBRSH and press Enter
   3) select MTART field click on Source filed(this is the field mapping) select MTART and press Enter
   4) select MAKTX field click on Source filed(this is the field mapping) select MAKTX and press Enter
   5) select MEINS field click on Source filed(this is the field mapping) select MEINS and press Enter
finally
save & come back
6) select radio-Button 6 and execute
   Maintain Fixed Values, Translations, User-Defined Routines
   Create FIXED VALUE Name & Description as MM01
   Create Translations Name & Description as MM01
   Create User-Defined Routines Name & Description as MM01
   after that delete all the above three just created in the 6th step
   FIXED VALUE --MM01
   Translations --MM01
   User-Defined Routines --MM01
   come back
7) select radio-Button 7 and execute
   Specify Files
   select On the PC (Frontend) -- and click on Create button(f5)
                                  give the path of the file like "c:\material_data.txt"
                                  description : -
                                  separators as select comma radiao- button
   and press enter   save & come back
8) select radio-Button 8 and execute
   Assign Files
   Save & come back
9) select radio-Button 9 and execute
   Read Files
   Execute
   come back
   come back
10) select radio-Button 10 and execute
    Display Imported Data
    Execute and press enter
    come back
    Come back
11) select radio-Button 11 and execute
    Convert Data
    Execute
    come back
    Come back
12) select radio-Button 12 and execute
    Display Converted Data
    Execute & come back
13) select radio-Button 13 and execute
    Start Direct Input Program
   select the Program
   select continue button
go with via physical file
give the lock mode as 'E'
and execute
Thanks & regards
sreeni

RH8 - Why do once-fixed auto-sizing, ranked index issues reappear in 64-bit build environment?

I'm using RoboHelp 8.0.2.208 fully integrated with Telelogic Synergy source control, and we generate WebHelp. In building the latest release (117) of our software the generated WebHelp has a couple of issues that weren't present in the previous builds. First, the auto-size pop-up topics are now cutting off text when they worked fine pre-117. Second, changes we'd made several releases ago to remove ranking in the index reverted, and we're back to ranked index results.
Our long-suffering Build Manager and I are trying to troubleshoot why this happened, and there are a couple of variables as of this build. Our developers are moving to PowerBuilder 12 as of this release, but we don't see anything that points to this being connected to the issue. Another variable is that we're in a new build environment that uses 64-bit OS. (There were the recent sunspots, too, but, naa, don't think that's it...). While I've found in this forum a discussion regarding the cut-off text in pop-ups (BTW, no extra lines ever had to be added), as well as the discussion about how to remove ranking in the index (http://forums.adobe.com/message/2901513#2901513), we'd like to find out what caused these issues to show up again now.
This also raises for us the issue of legacy files and how RoboHelp handles those in upgrading. It was when we moved from RoboHelp 5X to RoboHelp 8 that we first encounted and solved the ranked index issue. My recollection is that we also dealt with the pop-ups not autosizing properly at that time, although I haven't been able to find documentation about how we resolved that.
Is there code within our custom skin, old .css file, template/layout, or ? that could be bringing legacy issues into play here? What do the files in WebHelp5Ext do? Will migrated template_skin and/or template_stock eventually cause legacy issues? Does each subsequent RoboHelp release update to current coding standards?
Thanks!

CelesteD wrote:
1] Is there code within our custom skin, old .css file, template/layout, or ? that could be bringing legacy issues into play here?
2] What do the files in WebHelp5Ext do? Will migrated template_skin and/or template_stock eventually cause legacy issues?
3] Does each subsequent RoboHelp release update to current coding standards?
1] Code that is good now will likely have problems in the future. Code created in the days of DOS can now be a problem. In this forum we can only address identified issues.
2] Perhaps Willam can answer the first part. The second part has a similar answer to 1 above. Right now it should work, one day it might not but when and if that happens, I am sure Adobe will fix it.
3] If you mean the code that makes RoboHelp work, only Adobe would know that. If you mean the web outputs, I understand they meet current W3C standards but the standards are everchanging, all Adobe can do is meet the current standards and change things when the standards change.
Does that answer the questions?
See www.grainge.org for RoboHelp and Authoring tips
@petergrainge

Trex: Preparation failed: index operation

Hi gurus,
We have a Trex 7.0 server working on a seperate server. When I look at Index Administration, the statuses of the indexes are all green. But when i look at the indexing monitor, one of our indexes is in red status, It says Indexed:13342, Errors: 15610.
When I click on the Errors link, It says: "Trex: Preparation failed: index operation".
I searched the trace files but couldn't find sthg useful.
Any suggesstions?
Tolga

Hi Tolga,
please check your installation with the information mentioned by Jochen Goeeninger (long post) in the following thread Trex HTTP Server not seen in Portal
Furthermore try to open one of the documents you are trying to index via HTTP from the TREX Server.
With the result of that, I would suggest to open support message under http://service.sap.com/message
Hope this helps,
Robert

Search Indexing Issue

Hi All,
I am doing search indexing in ATG10. The following error is occurred while doing indexing .
Can you please give me your valuable suggestions to solve this error.
Content Item: atgrep:/CustomProductCatalog_production/book/prod117063?locale=el
atg.repository.search.indexing.IndexingException: java.lang.RuntimeException: CONTAINER:atg.repository.RepositoryException; SOURCE:org.jboss.util.NestedSQLException: Transaction
is not active: tx=TransactionImple < ac, BasicAction: 7f000001:c850:4ec0a480:60d81 status: ActionStatus.ABORTING >; - nested throwable: (javax.resource.ResourceException:
Transaction is not active: tx=TransactionImple < ac, BasicAction: 7f000001:c850:4ec0a480:60d81 status: ActionStatus.ABORTING >)
Regards,
Rams

Hi Rams,
Even we have faced the similar issue and we have done the following
1. Changed the default transaction timeout from 300 to 10000ms in /jboss/server/<server-name>/deploy/transaction-jboss-beans.xml
2. Increased the JBOSS connection pool size to min as 40 and max as 100
If you are unable to do indexing even after changing the above settings
3. Modify AEConfig.xml present in the bin folder of searhEngineDirectory with the below settings
MemoryIncrementSize - 0*00800000
MemoryReserveSize - 0*100000000
MemoryThreshold - 0*008000000
4. Even if you are facing indexing issue, modify the below property in Configuration component
/atg/search/routing/Configuration
defaultPartitionPhysicalCount = 1 - default value is 19, you can have it one or two
Hope the above informations helps you
Shabari

Trex is not indexing Outlook messages .msg

Hi Everybody,
Does anyone know why TREX is not indexing .msg (Outlook mail) files?
In [Help.sap.com|http://help.sap.com/saphelp_nw04s/helpdata/en/55/cb634114b78047e10000000a1550b0/frameset.htm] it says it should be supported...
Any ideas?
Thanks in advance,
Fede

Thanks Raghu!
You solved my problem!!
The definitive solution was:
In portal (SysAdm > SysConf > KM > CM > Utilities > MIME Types) I've added this settings:
Name "msg"
Default Extension ""
Extension "msg"
Locale ""
MIME Type "application/vnd.ms-outlook"
Resource Bundle File ""
Resource Key ""
In <disk_drive>:\usr\sap\trex_<instance_number>\TREXValidMimeType.ini, I've added the following rows:
application/vnd.ms-outlook
message/rfc822
Now TREX is indexing .msg outlook files!
Thanks Again,
Fede

Check Index Existence - Trex SOLMAN Solution Database Issue

Hi Experts,
I have done all the configuration part for having solution database functionality in SAP SOLMAN system. Our basis team has also installed TREX successfully.
But while checking in transaction CRMC_SAF_TOOL, everything is green except Check ndex Existence.
We are getting the below error messages:
Failed [Check=Index Existence ,Knowledge Base=SDB ,Language=E]
Failed [0 ]
Failed [No. of Documents=1-]
On discussion with Basis Team, I came to know that Indexes are getting created at putty level and while going to the indexes (Menu - Goto- Indexex (F8) in CRMC_SAF_TOOL, everything is seems to be fine and successful. No error there.
Please suggest where I am going wrong as I am still not able to search in IS02.
Thanks & Regards,
Madhu

Hello,
Sometimes this error is seen when an older version of Trex being used.
Are you on a current version of Trex? If not you may want to consider upgrading to a current version?
Additionally, sometimes documents saved via content management cannot be indexed,
as no url can be determined for them.
4. knowledgebase SDB_ATTACHMENT:
   The URL's do not get determined and this is why nothing would be
   put to TREX.
   Herefore, please follow the following note:
   685521 Logon data for /sap/bc/contentserver service
This may explain the HTTP:/no. DOcuments =1- error ???
So please ensure you have set the URL for the content server as per Note 685521
Regards,
Paul

TREX Indexing issue. Form based publishing XML files.

In order to test my TREX indexing functionality, I have created a lot of test XML files (with form based publishing) containg test words.
What I experiments is that TREX only index the words that is presented on the 'Show' form of the xml document. The template created in the XML Forms Builder that is used to present the document. But thing is, that I only want to index the words from the XML source file, not from the 'Show' form template. Because on that i only have a static label with constant text.
How do I set TREX to index the source XML file and not the show form template representation of the file?

Hi Flemming,
this works as designed. If you want index the whole XML data, you have to store your XMLForms items in a repository on which no "XMLFormsResourceFilter" is defined.
But if you do that, all the items are rendered as XML and not as HTML. So the SearchResult only gives you links pointing to XML files.
The only way to index more informations (from my point of view) is to store your additional information into ResourceProperties (you have to design your XMLFormsProject). These properties can be found by the search, if they are set to "indexable=yes".
Hope this helps
--Matthias

TREX unable to index only some files

Hi All!
I've got problem with my TREX (6.1 SP14 running on portal maschine SAP EP 6.0 SP14).
I've created some repositories, physically residing on that server's filesystem. It's an WebDAV repository and two CM (FSDB and DBFS).
Created an Index.
It indexes those repositories, and after that I see, that it can't index content of some files, while normally indexes others. It can't index *.doc, *.xls, *.pdf, *.ppt. It normally indexes *.txt, *.zip and folders.
In application log (in portal) I can see, that error code is 401 (6401), 'not authorized'.
If I open (for example) some *.doc file properties and try to access 'Settings'>'Classification', I recieve an error:
"An error occurred during classificationError while classifying resource /FSDB_local/testdocFSDB.doc for taxonomy localfiles_tax.com.sapportals.wcm.WcmException: HTTP Status Code 401: Unauthorized (Errorcode 6401)".
In IIS log there is:
"2006-01-26 17:51:01 W3SVC1 10.2.1.67 PROPFIND /localfiles/EPtest/Logging.doc - 80 GMCS\MYUSER 10.x.x.x SAPCMWebDAVRepository1.1 $Version=0;+MYSAPSSO2=AjExMDAgABJwb3J0YWw6 207 0 0"
In security log of windows system I can only see 'successful' events like :
"Event Type: Success Audit
Event Source: Security
Event Category: Object Access
Event ID: 560
Date: 26.01.2006
Time: 20:41:13
User: HOSTNAME\SAPServiceSID
Computer: HOSTNAME
Description:
Object Open:
 Object Server: Security
 Object Type: File
 Object Name: D:\repositories\FSDB
 Handle ID: 13960
 Operation ID: {0,1066813}
 Process ID: 924
 Image File Name: D:\usr\sap\G03\JC00\j2ee\os_libs\jlaunch.exe
 Primary User Name: SAPServiceSID
 Primary Domain: HOSTNAME
 Primary Logon ID: (0x0,0xEFAE)
 Client User Name: -
 Client Domain: -
 Client Logon ID: -
 Accesses: SYNCHRONIZE
 ReadAttributes
 Privileges: -
 Restricted Sid Count: 0
 Access Mask: 0x100080
For more information, see Help and Support Center at http://go.microsoft.com/fwlink/events.asp.
What could it be?
All permissions (on portal side and on filesystem) are the same for all files.
Thanks in advance for help and support!
Regards,
DK

I have a similar issue with a web repository.
I have an alternative URL set in the URL generator and I also do not have a locked index_service user (password is also reset) but I am getting the same 401 (6401 TREX error) problems as others.
Anybody know how to solve this problem?
How does TREX authenticate with the portal using the index_service user? How does TREX know what credentials to use? If it is ticket based authentication, how is this ticket set in the TREX system?
Thanks,
Tom
http://tomparle.com/
http://deadlytech.co.uk/

Trex failing to index word docs in FS repository

Hi,
I am having errors while indexing a FS repository.
The errors in TREX monitor are
return message:Content-Length -vs- Actual Read mismatch
return code:8030
Document Status: Preparation Failed
I tried reindexing these failed entries, but did not work out.
All of these are word docs. Moreover Crawler errors for
folders are also visible in the errors list.
Looking into the TREX trace , was able to get the follwing.
HTTP-GET failed for URL <............>with Errorcode -30 , but HTTP-HEAD worked, trying again,
Mimetype application/msword is not based on TEXT, but was detected as text type; content might be corrupted and will be ignored
When I tried opening some of these docs from Windows,
I did not have access to the same itself and got an error. Would this also be the case for the index_service user?
Also not all docs which are visible on the Windows side are visible from the portal in the KM nav iview.
Platform :EP 7.0 SP13, TREX 7.0
Any help would be appreciated
Rgds

Hello,
A problem related to the search for this FS repository is that search takes a very long time to display results.
The datasource is quite big with nearly 90,000 docs and the index is supposed to index external links too with
indexContentOfExternalLink, indexContentOfExternalLink properties set also.The search scope is based on Indexes in the search iview .
When a normal user who has the relevant role for the searh iview runs a search it takes a very long time, nearly 30 min!!
But if a super admin runs the same , it comes up with results immediately.
Is this some kind of authorisation issue.The index is having everyone full control and I was not able to see any thing much in the default trace too.
Is there any particular trace/log file to be checked for this?Has something been missed out in the index creation process?
Hope someone can comment on this
Rgds

Re indexing issues in Apple Loop Browser (LP 9.0.2 / OSX 10.6.1 / Digi 002r

I'm trying to re index my loop browser in LP9 / OSX10.6.1
Are you having the same problem?
After trashing the 'Apple Loop Index', I pull in a folder from an external drive and notice that previous folders added to my browser have gone, including Jam Packs.
Also, I can't find the 're-index' option within the loop browser search option (right click brings down a list but the re-index option isn't there)
Interested to see whet you find.

Yep, I also had issues, where indexing the LP8 way (dragging folders with loops from the Finder onto the Logic loop browser) just bluffed to work (I dropped the folder on the LB, a progress window appeared, things apparently got done, but then nothing, no new entries in the Loop Browser.
So it seems that it is a L9 problem, since our Macs' are different in every aspect (Me: PPC- you: Intel, me: Leopard- you: Snow Leopard).
I finally decided to put all my loops in their default install locations and reindex them from there:
*Library/Audio/Apple Loops/Apple:*
Jam Pack 4 - Symphony Orchestra
Jam Pack Voices
Jam Pack World Music
*Library/Audio/Apple Loops/iLife Sound Effects:*
(13 folders)
*Library/Audio/Apple Loops/User Loops:*
(some folders with my own and 3rd party Loops)
*Library/Application Support/Garageband/Apple Loops:*
Apple Loops for Garageband
Jam Pack 3- Rythm Section
for thorough reindexing, here are the steps:
1. Quit Logic and/or any other app that uses Apple Loops.
2. Go to all the locations I pointed out and trash all the index files from the *Apple Loops Index* folders. Also check the same locations in your Users/'you'/Library folder.
3. Start Logic with a new empty project, open the loop browser window by hitting the o key. Now drag the folders mentioned above in bold onto your loop browser. Logic should now index them correctly, including the foldermenu !http://farm3.static.flickr.com/2743/4085966954bf4fc039d7o.png! in the loop browser.
regards, Erik.

LiveCache - LC10 message - Index issue

hi,
LiveCache - LC10, - Problem Analysis ->Performance -> Database Analyser -> bottleneck report
The message read as follows:
LiveCache- Bottle-neck messages:
2 tables contain > 1.000.000 records but only 20.000 rows will be sampled for statistics.
Table SAPR3./SAPAPO/ORDKEY contains 8247892 rows(205921 pages), sample rows 20000
Table SAPR3./SAPAPO/STOCKANC contains 1319385 rows(25413 pages), sample rows 20000
Looks this would affect the processing and the delay may be an issue for CIF queue processing.
Wondering, if sampling could be increased ? is there any oss note available to do this. This index is implemented using BADI
Any input on this issue is appreciated.
Thanks,
RajS

Please check the following notes regarding changing the sample of the statistics run:
Note 808060 - Changing the estimated values for Update Statistics
Note 927882 - FAQ: SAP MaxDB Update Statistics
Kind regards,
Mark

Trex 7.0 index issue

Similar Messages

Maybe you are looking for