Regarding Performance Issue in MI Application

Hii All,
We have made a SmartSync MI Application.Everything thing is working fine.But the scenario is this that when we try to navigate through the application from one page to another ,for example if i click on one link to get details corresponding to that link(i.e) we are retreiving the data from the local database, but this is taking a lot of time and according to our buisness need we have to reduce this time to fetch the data.
Can anyone help me regarding this issue what can be done either in coding part, what parameters & standards to follow to improve the performance.
Thanks,
Devendra Phate

hello deven,
if you are only focusing on your application's performance
on the aspect of usability i.e. less waiting time, fast
response UI, i would personally suggest you to use AJAX.
put some processing part on the browser side and making
data retrieval from MI asynchronous... this way, your user
don't have to wait for all data to be presented, but rather
could work on the data presented in pieces (i.e. asynchronously).
anyway try googling for AJAX...
regards
jo

Similar Messages

Performance Issue - 32 Bit Applications vs 64 Bit DB Server

Dear All,
My Current set up is as follows:
Operating System: Windows 2003 Standard Edition R2 64 Bit
Database: Oracle 10.2.0.1 64 Bit
Client Connectivity: Oracle 10.2.0.1 32 Bit
Application: Lotus Domino 7.0.2 FP3 32 Bit
In the above set up,
When Application try to fetch the records from DB through client, it takes so much of time as compared to when i run the same with (OS & DB in 32 bit).
kindly let me know how to do workaround on this, analysation have to be made to overcome this performance issue.
thanks & regards
suresh

1. Check for available statistics.
2. Verify plans against 32 bit and 64 bit databases.
3. Analyze the sql trace for the session.

Performance issue on swing application

Hi,
I have one swing application and i am facing one performance issue after minimizing the applicaton. The same application is taking 2-5 seconds to redraw the screen again when i am trying to maximize it. Is there any way to improve this redraw delay ?..Is there any way to tweak the Event Dispatcher thread and by this we can overcome correct me if i am wrong? .
Is there anybody facing such kind of issue please help me...
Thanks in Advance,
Anish.

Encephalopathic wrote:
good guess. What do you think? Loading the images in the painting methods?That is a common mistake. To the OP, some common errors when overriding paintComponent include:
1. Loading an image each time paintComponent() is called. Just load it once, cache it, and only draw it in paintComponent, this saves you from the IO of loading the image on repaints.
2. Doing application logic/complex operations in paintComponent. paintComponent should be kept fast, only do painting operations, keep long-running tasks, IO, etc. out of it.
3. To a lesser extent, doing lots of complex painting on a repaint call, maybe lots of transforms, translucency, xor-ing, whatever. For most common painting (at least in my work, YMMV), this isn't a problem, but if it is you should consider drawing all "static" stuff to a buffer (such as a BufferedImage) and just painting the BufferedImage, as opposed to drawing all the static stuff each paintComponent call.
The idea is just to keep the painting logic as fast as possible, to avoid slow repaints as you're seeing. Sometimes people don't realize that paintComponent() will be called every time your component needs to be even partially repainted (minimized/maximized window, partially hidden by another window then revealed again, stretched by window resizing, etc.). That's why things like IO are out of the question.
Of course, I'm babbling now. This may not even be your problem, or you may already know all this stuff. If so, just ignore this post!

Regarding performance issue in time dependent hierarchie.

hai
we are loading time dependent hierarchies from flatfile to bw,it is weekely load,we have nearly one million records loaded,we got an issues regarding change of these hierarchies timely.when ever there occurs a change in hierarchy a new row is added to the table,it is degrading the performance,can any one of you please suggest how to over come performance related issues regarding time dependent hierarchies.
Regards
Srinivas.G

hello deven,
if you are only focusing on your application's performance
on the aspect of usability i.e. less waiting time, fast
response UI, i would personally suggest you to use AJAX.
put some processing part on the browser side and making
data retrieval from MI asynchronous... this way, your user
don't have to wait for all data to be presented, but rather
could work on the data presented in pieces (i.e. asynchronously).
anyway try googling for AJAX...
regards
jo

Performance Issue in Web Application: Event 12605 "Web: Text Content"

Dear All,
we are about to deploy a couple of rather complex web applications (built with WAD) and are currently running performance tests with Mercury Loadrunner.
According to these tests our web application become very slow once more than 30 (virtual) users are running the application.
When analyzing the BW statistics data we found out that most of the runtime of our application comes from an event called "Web Java: Return Text-Type Content" (ID 12605).
Has anyone experienced the same issue? Or does anyone know what exactly this event means?
We are using XPATH in our web applications. Could this be a cause for poor performance?
Any help is highly appreciated!
Best Regards
Christian

Can you tell me more about the detail of loadrunner stress test for BI WAD scripts? We met error when run the load runner for BI WAD report.
1. We develop the BI report basing WAD 7.x, setup Iview in sap EP for BI report and get the BI report URL.
2. We fill the report URL in IE(6.0) to run this report and display the bi report successfully ,
eg: http://epserver:50000/irj/servlet/prt/portal/prtroot/com.sap.ip.bi.web.portal.integration.launcher?sap-bw-iViewID=pcd%3aportal_content%2fcom.ahepc.BI%2fcom.ahepc.iView%2fcom.ahepc.FI%2fcom.ahepc.BI_I_ZWAFI_006&sap-ext-sid=1

Performance Issue Of Java Application on Pentium 4

Dear All
I have finished an application program, which runs smoothly in PC with Intel Pentium 3 Processor with high performance, but the performance is highly decreased with Intel Pentium 4 Processor. Is it a limitation of Java?
Im using MS Access database.
I have restructured the queries but the performance is highly enhanced in P3 but it degrades and sometimes there is no consistency in the performance timmings in P4.
System Configuration are:
P3 : 256 MB RAM with Windows 2000 SP2,MS ACCESS 2000
P4: 256 MB RAM with WIN 2000 SP2,MS ACCESS 2000
Is there any solution to this problem?
Pls help!
Best Regards
Sourav

256MB of RAM for a Windows 2000 machine?!?
You're probably swapping memory.
- Go into Task Manager and see what is running on each machine.
...and see if you can not start up some things.
- See how much memory you have Available on each BEFORE you run the app.
- use java -Xmx option to specify less than the available memory you have.

Regarding Performance Issue

Hi Friends,
i have a FI report, Main Cash Vocher Printing, the report is very slow so to check its performance i have done the SQL Trace in ST05.
in SQL trace i have found all the DB tables used and time taken to execute the select statement.... in that list i also found and Cluster Table RFBLG , but i did not used these cluster table in the program.
and i also found more than 500 entries of Obj.NameRFBLG in that ...
Though its duration is small , ,, can we control that so that my report performance van be increased
its Urgent...
Useful reply will be awarded maximum
Regards
Kumat M

Hi Kumat,
RFBLG is the Cluster table of the logical table BSEG. That means the table BSEG stores its rows in the database table RFBLG. If the access to RFBLG is fast enough, you could ignore those entries and concentrate on red lines in the trace.
Regards
Ralph Ganszky

Regarding Performance issues

Can anybody share pure BI Performance document of DB02, St02,RSRST analysis does purpose.
Can anyone share it., Thanks in advance.

HI,
check below links
http://www.sap-press.de/download/dateien/1049/sappress_bw_performance_optimization_guide_080.pdf
http://wiki.sdn.sap.com/wiki/display/BI/Aggregates--SAPBWQueryPerformance
http://sap.seo-gym.com/performance%20tuning%20for%20queries.pdf
Click on search button in SDN and Google you can find lots of docs
Regards
KP

Performance issue in web application

We have installed a web application that uploads a XML file to the server. The application is still in its nascent stage but deployed.
After analysis we found out that 50 to 100 hits take place at any point of time. We store the XML file as a blog in oracle and use Xquery to retrieve the data. The core apis used are Java Servlets and JSPs and the web server used in Apache Tomcat.
The hits for upload will increse by 500-600 at point of time.
I would like to know the following.
Is the technology used sufficient to handle the issue?
How can i handle the load balance in the server?

Use grinder or some other test tool to find out.

Regarding performance issue in SELECT - need suggestions

Moderator message - use a more informative subject in future - having two or more posts with the same subject isn't helpful
Hi All could you plese send me the alternative code for this select quesy:
form read_data_from_GIPD.
select * from zbacks_gipd
    where failf = '2'
      and recno in recno.
    clear tab.
    clear result_total. "reset the record counter
    tab-werks = zbacks_gipd-WERKS.         "plant
    tab-aufnr = zbacks_gipd-AUFNR.         "production order no.
    tab-matnr = zbacks_gipd-MATNR.         "material
    tab-erdat = zbacks_gipd-BUDAT.         "transaction date
    tab-recno = zbacks_gipd-recno.
    TAB-MENGE = zbacks_gipd-menge.         "qty
    TAB-USAFL = zbacks_gipd-usafl.         "usage/consumption ind.
*          TAB-EXPCD            "Expenses code
    TAB-SAKNR = zbacks_gipd-saknr.         "G/L account number
    TAB-IDNUM = zbacks_gipd-idnum.         "BACKS identification no.
    TAB-KUNNR = zbacks_gipd-kunnr.         "Customer Number
    TAB-USACD = zbacks_gipd-usacd.         "usage code
    TAB-ORDCD = zbacks_gipd-ordcd.         "order code
    TAB-LOSTY = zbacks_gipd-losty.         "loss type
    TAB-COTAX = zbacks_gipd-cotax.         "consumption tax
    TAB-MWSKZ = zbacks_gipd-mwskz.         "tax code
    tab-runum = zbacks_gipd-runum.         "running number
    tab-failf = zbacks_gipd-failf.     "failure flag
    tab-msgnr = zbacks_gipd-failf.     "failure flag
    tab-matdoc = zbacks_gipd-docm1.    "material doc
* BEGIN OF SR#20293 WITH DEVK955638
*    tab-msgtx = 'Repost 2nd step'.
    tab-msgtx = 'Repost 2nd step'(057).
*   END OF SR#20293 WITH DEVK955638
*    tab-erdat_raw = tab-erdat. "keep the MM/DD/YYYY format
*    CONCATENATE TAB-ERDAT+6(4) TAB-ERDAT+0(2) TAB-ERDAT+3(2)
*           INTO TAB-ERDAT.
     concatenate tab-erdat+4(2) '/'
                 tab-erdat+6(2) '/'
                 tab-erdat+0(4)
               into tab-erdat_raw.
*Process only a particular set of usage indicator first
    if tab-usafl = p_usafl.
      APPEND TAB.
      result_total = result_total + 1. "Increase the record counter
    endif.
endselect.
* BEGIN OF SR#20293 WITH DEVK955638
* write: / 'Re-processing records for usage indicator', p_usafl.
write: / 'Re-processing records for usage indicator'(056), p_usafl.
* END OF SR#20293 WITH DEVK955638
endform.
Note : how can i avoid the select endselect statement , here he move data from database table to internal table .
Edited by: Julius Bussche on Jan 19, 2009 1:52 PM
Please use code tags
Edited by: Matt on Jan 19, 2009 4:46 PM

Hi,
     Declare an Internal table with zbacks_gipd type,
     Move the records into that table based on conditions,
     Then loop that internal table, and fill tab internal table.
form read_data_from_GIPD.
Types : begin of ty_zbacks_gipd.
             include structure zbacks_gipd.
Types : end of ty_zbacks_gipd.
Data : it_zbacks_gipd type table of ty_zbacks_gipd,
             wa_zbacks_gipd type ty_zbacks_gipd.
select * from zbacks_gipd into table it_zbacks_gipd
       where failf = '2'
      and recno in recno.
loop at it_zbacks_gipd into wa_zbacks_gipd.
clear tab.
    clear result_total. "reset the record counter
    tab-werks = wa_zbacks_gipd-WERKS.         "plant
    tab-aufnr = wa_zbacks_gipd-AUFNR.         "production order no.
    tab-matnr = wa_zbacks_gipd-MATNR.         "material
    tab-erdat = wa_zbacks_gipd-BUDAT.         "transaction date
    tab-recno = wa_zbacks_gipd-recno.
    TAB-MENGE = wa_zbacks_gipd-menge.         "qty
    TAB-USAFL = wa_zbacks_gipd-usafl.         "usage/consumption ind.
*          TAB-EXPCD            "Expenses code
    TAB-SAKNR = wa_zbacks_gipd-saknr.         "G/L account number
    TAB-IDNUM = wa_zbacks_gipd-idnum.         "BACKS identification no.
    TAB-KUNNR = wa_zbacks_gipd-kunnr.         "Customer Number
    TAB-USACD = wa_zbacks_gipd-usacd.         "usage code
    TAB-ORDCD = wa_zbacks_gipd-ordcd.         "order code
    TAB-LOSTY = wa_zbacks_gipd-losty.         "loss type
    TAB-COTAX = wa_zbacks_gipd-cotax.         "consumption tax
    TAB-MWSKZ = wa_zbacks_gipd-mwskz.         "tax code
    tab-runum = wa_zbacks_gipd-runum.         "running number
    tab-failf = wa_zbacks_gipd-failf.     "failure flag
    tab-msgnr = wa_zbacks_gipd-failf.     "failure flag
    tab-matdoc = wa_zbacks_gipd-docm1.    "material doc
* BEGIN OF SR#20293 WITH DEVK955638
*    tab-msgtx = 'Repost 2nd step'.
    tab-msgtx = 'Repost 2nd step'(057).
*   END OF SR#20293 WITH DEVK955638
*    tab-erdat_raw = tab-erdat. "keep the MM/DD/YYYY format
*    CONCATENATE TAB-ERDAT+6(4) TAB-ERDAT+0(2) TAB-ERDAT+3(2)
*           INTO TAB-ERDAT.
     concatenate tab-erdat+4(2) '/'
                 tab-erdat+6(2) '/'
                 tab-erdat+0(4)
               into tab-erdat_raw.
*Process only a particular set of usage indicator first
    if tab-usafl = p_usafl.
      APPEND TAB.
      result_total = result_total + 1. "Increase the record counter
    endif.
endloop.
endform.
Regards
Bala Krishna
Edited by: Bala Krishna on Jan 19, 2009 6:48 PM

Performance issues with class loader on Windows server

We are observing some performance issues in our application. We are Using weblogic 11g with Java6 on a windows 2003 server
The thread dumps indicate many threads are waiting in queue for the native file methods:
"[ACTIVE] ExecuteThread: '106' for queue: 'weblogic.kernel.Default (self-tuning)'" RUNNABLE
     java.io.WinNTFileSystem.getBooleanAttributes(Native Method)
     java.io.File.exists(Unknown Source)
     weblogic.utils.classloaders.ClasspathClassFinder.getFileSource(ClasspathClassFinder.java:398)
     weblogic.utils.classloaders.ClasspathClassFinder.getSourcesInternal(ClasspathClassFinder.java:347)
     weblogic.utils.classloaders.ClasspathClassFinder.getSource(ClasspathClassFinder.java:316)
     weblogic.application.io.ManifestFinder.getSource(ManifestFinder.java:75)
     weblogic.utils.classloaders.MultiClassFinder.getSource(MultiClassFinder.java:67)
     weblogic.application.utils.CompositeWebAppFinder.getSource(CompositeWebAppFinder.java:71)
     weblogic.utils.classloaders.MultiClassFinder.getSource(MultiClassFinder.java:67)
     weblogic.utils.classloaders.MultiClassFinder.getSource(MultiClassFinder.java:67)
     weblogic.utils.classloaders.CodeGenClassFinder.getSource(CodeGenClassFinder.java:33)
     weblogic.utils.classloaders.GenericClassLoader.findResource(GenericClassLoader.java:210)
     weblogic.utils.classloaders.GenericClassLoader.getResourceInternal(GenericClassLoader.java:160)
     weblogic.utils.classloaders.GenericClassLoader.getResource(GenericClassLoader.java:182)
     java.lang.ClassLoader.getResourceAsStream(Unknown Source)
     javax.xml.parsers.SecuritySupport$4.run(Unknown Source)
     java.security.AccessController.doPrivileged(Native Method)
     javax.xml.parsers.SecuritySupport.getResourceAsStream(Unknown Source)
     javax.xml.parsers.FactoryFinder.findJarServiceProvider(Unknown Source)
     javax.xml.parsers.FactoryFinder.find(Unknown Source)
     javax.xml.parsers.DocumentBuilderFactory.newInstance(Unknown Source)
     org.ajax4jsf.context.ResponseWriterContentHandler.<init>(ResponseWriterContentHandler.java:48)
     org.ajax4jsf.context.ViewResources$HeadResponseWriter.<init>(ViewResources.java:259)
     org.ajax4jsf.context.ViewResources.processHeadResources(ViewResources.java:445)
     org.ajax4jsf.application.AjaxViewHandler.renderView(AjaxViewHandler.java:193)
     org.apache.myfaces.lifecycle.RenderResponseExecutor.execute(RenderResponseExecutor.java:41)
     org.apache.myfaces.lifecycle.LifecycleImpl.render(LifecycleImpl.java:140)
On googling this seems to be an issue with java file handling on windows servers and I couldn't find a solution yet. Any recommendation or pointer is appreciated

Hi shubhu,
I just analyzed your partial Thread Dump data, the problem is that the ajax4jsf framework ResponseWriterContentHandler triggers internally a new instance of the DocumentBuilderFactory; every time; triggering heavy IO contention because of Class loader / JAR file search operations.
Too many of these IO operations under heavy load will create excessive contention and severe performance degradation; regardless of the OS you are running your JVM on.
Please review the link below and see if this is related to your problem.. This is a known issue in JBOSS JIRA when using RichFaces / ajaxJSF.
https://issues.jboss.org/browse/JBPAPP-6166
Regards,
P-H
http://javaeesupportpatterns.blogspot.com/

Performance issue session or bean?

hi,
i have a doubt regarding performance issue.
actually in my application, five jsp' s need to communicate each other. They share around 15 String Objects.
1. So right now i am using Session object to store
these 15 String objects.
2. Also my application can declare a class to
store all these 15 String objects as attributes and
then use <jsp:usebean> tag for communication
with session scope.
So now we have two methods for sharing information
in jsp application as stated above.
Can anyone help me which is the best method to follow as given above and why. Which does increase or decrease the web server load.
Thanks in advance.
Have a nice day.
Bye
Suresh

I'd go for wrapping up the data in a single object...not for any performance issues, just because it's far tidier.
As for performance, I doubt there is much difference although by wrapping up the data in one object you only have to call the session object once to get a handle on all 15 strings.

Regarding performance optimization and tuning...

hi all,
<b>please provide me the performance tuning scenarios and parameters of an R/3 system with Oracle..</b>
i heartly welcome all docs and pdf links or notes related to this issue..
please provide ur suggestions at the earliest...
expecting ur response..
<i>Vineeth</i>

Hello,
there are many SAP Notes regarding performance issues. Here are just a couple of them:
618868
805934
793113
805934
Please also have a look at the lists of the relating Notes at the end of each Note.
But still much more effective would be to read the book of
<a href="http://www.sap-press.de/katalog/buecher/titel/gp/titelID-1155?GalileoSession=66220888A2.lRCIISlE">T.Schneider Performance Optimization Guide</a>.
It's the best performance tuning guide. The course ADM315 (or BC315?) ist also very helpful.
Regards,
Natalia

BPC 10.0 Input Schedule refresh : investigating performance issues on frontend side

Hello,
We are running SAP BPC NW 10.0 SPS8.
We recieve a lot of complaints regarding performance issues.
More specifically BPC Input Schedule refresh takes ages (11 minutes actually)
I am a SAP Basis Administartor with some solid BW experience, but very linited BPC knowledge.
So raf we have tried the following :
* BPC Full and light ptmimize of the BPC Cubes.
* Updating Oracle Statistics
* Activating ABAP/SQL traces*
* Checking USJTAT Logs
Then it rutned out that the perfromance problems werre not caused by the backend SAP BW server (only 9 seconds of bakend processing are displayed in UJSTAT).
The overwhelming majority of the processing is spent on the frontend side : on the computer of the user.
We really would like to know what is happening on BPC Frontend side, but are really lacking any investigative method .....
other that checking memory/CPU usage o nthe frontend computer.
From your experience, is there any tool/method to allow to conduct some investigation on what is going on the frontend side, when a BPC Input Schedule is refreshed ?
Thank you for your attention.

Hello Vadim,
We still need to get confirmation from the external consultants who developed the BPC solution, but as far as I can understand it :
We do not use use member formulas
We are indeed talking about EPM (without EPMRetrieveData functions)
The perfromance problems occur in DEV and QAS , after a restore of the BPC environment (backup from PRD)
We already had the problem in the past, but we never knew precisely how to solve it, it is as if there was some missing links between the restored BPC environment from PRD and the DEV/QAS system
One additional information, we checked the EPM CLient Log , and here are the errors displayed while refreshing the input schedule :
The thing is that I have no idea what it means ...
?FILE_TYPE:DAAA96DE-B0FB-4c6e-AF7B-A445F5BF9BE2
RECORD_SEPARATOR:35
COLUMN_SEPARATOR:124
COLUMNS:Time|Severity|Category|Method|User|DSRRootContextID|DSRTransactionID|DSRConnectionID|DSRCounter|Thread|Text
SEVERITY_MAP:DEBUG|Debug|INFO|Information|VERBOSE|Verbose|WARNING|Warning|ERROR|Error|FATAL|Fatal
HEADER_END
2014-04-18 08:41:17,446|ERROR|ReportManagement|FPMXLClient.Office.EPMExcelReportManagerRepository.GetReportManager|BPCUSER|||||1|Unable to get the Object property of the OLEObject class#
2014-04-18 08:41:18,613|ERROR|ReportManagement|FPMXLClient.Office.EPMExcelReportManagerRepository.GetReportManager|BPCUSER|||||1|Unable to get the Object property of the OLEObject class#
2014-04-18 08:41:18,630|ERROR|ReportManagement|FPMXLClient.Office.EPMExcelReportManagerRepository.GetReportManager|BPCUSER|||||1|Unable to get the Object property of the OLEObject class#
2014-04-18 08:41:18,648|ERROR|ReportManagement|FPMXLClient.Office.EPMExcelReportManagerRepository.GetReportManager|BPCUSER|||||1|Unable to get the Object property of the OLEObject class#
2014-04-18 08:41:18,731|ERROR|ReportManagement|FPMXLClient.Office.EPMExcelReportManagerRepository.GetReportManager|BPCUSER|||||1|Unable to get the Object property of the OLEObject class#
2014-04-18 08:41:37,171|ERROR|ReportManagement|FPMXLClient.ReportManagement.DynamicReport.Load|BPCUSER|||||1|bytes = 1#
2014-04-18 08:41:39,348|ERROR|ReportManagement|FPMXLClient.ReportManagement.DynamicReport.Load|BPCUSER|||||1|bytes = 0#
2014-04-18 08:41:39,474|ERROR|ReportManagement|FPMXLClient.ReportManagement.DynamicReport.Load|BPCUSER|||||1|bytes = 0#
2014-04-18 08:41:39,536|ERROR|ReportManagement|FPMXLClient.ReportManagement.DynamicReport.Load|BPCUSER|||||1|bytes = 0#
2014-04-18 08:41:39,597|ERROR|ReportManagement|FPMXLClient.ReportManagement.DynamicReport.Load|BPCUSER|||||1|bytes = 0#
2014-04-18 09:00:22,088|ERROR|Dataservice|FPMXLClient.BPCProxy.MetaData_Data.DataServiceHelper.ExecuteReportAsStream|BPCUSER|||||ServerCallThread_666f378d-d824-4a8a-bf38-0da8493e2382|An error has occured on the server#
2014-04-18 09:00:22,252|ERROR|BPC10Connection|FPMXLClient.Connection.RESTConnection.ExecuteReport|BPCUSER|||||ServerCallThread_666f378d-d824-4a8a-bf38-0da8493e2382|An error has occured on the server#
2014-04-18 09:00:22,268|ERROR|OfficeWrappers|FPMXLClient.Office.EPMReportManager.ExecuteReport|BPCUSER|||||1|An error has occured on the server#
2014-04-18 09:00:22,317|ERROR|OfficeWrappers|FPMXLClient.Office.EPMExcelReportManager.RefreshReports|BPCUSER|||||1|An error has occured on the server#
2014-04-18 09:00:22,447|ERROR|AnalyzerOne|FPMXLClient.AddIn.RefreshSheetWithScan|BPCUSER|||||1|Execution of Report Default Report failed : #
2014-04-18 09:00:22,537|ERROR|Automation|FPMXLClient.EPMAddInAutomation.RefreshActiveSheet|BPCUSER|||||1|Execution of Report Default Report failed : #

PrepareLogin Performance Issue

Hi Experts,
Can you please help me know what are the things I should consider
because I have performance issue
when my application is executing "preparelogin.do" it is taking about 30 minutes:
https://myserver:50001/myappb2b/user/preparelogin.do?secure=ON
I have the followng config.xml:
<action path="/preparelogin" type="com.sap.isa.user.action.PrepareLoginBaseAction">
     <forward name="success" path="/login.do"/>
     <forward name="failure" path="/logon/login.jsp"/>
     <forward name="darklogin" path="/login.do"/>
     <forward name="umelogin" path="/base/empty.jsp"/>
</action>
<action path="/login" type="com.sap.isa.user.action.LoginBaseAction">
     <forward name="success" path="/toModule.do?prefix=&page=/postLogin.do"/>
     <forward name="failure" path="/logon/login.jsp"/>
     <forward name="pwchange" path="/logon/pwchange.jsp"/>
     <forward name="failureUME" path="/base/error_ume.jsp"/>
</action>

<action path="/toModule" type="org.apache.struts.actions.SwitchAction"/>

<action path="/postLogin" forward="/b2b/postlogin.do"/>
<action path="/b2b/postlogin" type="com.sap.isa.isacore.action.b2b.PostLoginAction">
<forward name="success" path="/b2b/campaignfromMIG.do"/>
</action>
Can you please give me idea what may be the possible cause of login performance when it executes PrepareLoginBaseAction? Or are there configurations I need to look at?
My application uses LDAP authentication then goes to SAP UME authentication.
Thank you very much in advance.
jemaru

Check some of my notes here:
[http://wiki.colar.net/sap_crm5_0_isa_trex_performance_settings]
In particular check the JDBC connection(this cause issue very similar to yours) and gateway connection sections, those can cause serious issues like this, this default settings use to be seriously too low, until recent releases.
A problem in production is harder to debug / trace, but if you can, do the single activity trace as somebody suggested.
Another option is to increase the logging, that will give you infos on how long each RFC / action class take and should allow you to locate what's slow.
When the logs are on, they are huge and it's hard to find the relevant data, so i have a small script here, that helps finding the interesting stuff (after turning on logging, you run the script and it will print out interesting lines found in the logs).
See: [http://wiki.colar.net/isa_action_rfc_tracer] ... you can use that script, or alternatively use tail/grep to do the same thing (if on *nix)
Note: edit the script, set the correct SID for your system and set showRFC to true.
compile it (javac) and run it (java)
More infos on tracing / debugging options:
[http://wiki.colar.net/b2c_customization]
Edited by: Thibaut Colar on Oct 12, 2010 10:37 AM

Regarding Performance Issue in MI Application

Similar Messages

Maybe you are looking for