SLD Landscape Question

Question

Hi
We are about to implement the SLD landscape for our PI systems.
From the documentation (SDN and HELP.SAP.COM), it is recommended to have all SLD Data Suppliers send to the PI Production SLD Bridge.  The SLD data is then automatically forwarded to the DEV/QA SLD Bridge (Design Time SLD).
This is straight forward and I can see how to get the ABAP stack to look at the correct SLD Server.
However, when it comes to the JAVA stack, I can't see how to configure the SLD CLIENT of an instance in the design-time landscape to use the design-time SLD Server.
Version of Netweaver in 7.0.
Thanks
Doug

Answer 1

SAP Web AS (Java) 6.30/6.40 Start-up: JProbe's Direct Support
This feature is available starting with JProbe 5.2.2. For details of application server integration, refer to the JProbe documentation. Here, the single steps for the SAP application server are briefly described.
Step 1: Create an Integration for the instance to be profiled.
      Click Tools -> Application Server Integration.
      Choose SAP Application Server 6.30.
      Click Create.
      To fill out the columns, follow the help you get via tooltip.
      Specify whether server or dispatcher should be profiled.
Step 2: Create a Configuration.
      Click Tools -> Manage J2EE Configurations.
      Click Add.
      Fill in Configuration Name and Integration.
      Optionally, you could specify an application to be profiled; this is just used to help to set the correct filters for profiling. If you know which packages or classes should be profiled, you can leave these fields empty; otherwise, specify the directory where the application is deployed and the .ear or .war file in this directory.
Step 3: Create the J2EE Settings.
      Click Session -> New J2EE Settings.
      Specify the configuration.
      Specify the other options, especially set filters (optionally, you could use the application specified in the configuration for filtering).
Step 4: Use the Connection Manager.
      Different to "startup via jlaunch" below the Connection Manager has to be used.
      Click Tools -> Options -> Advanced Session; mark "Use Connection Manager."
Step 5: Check the property 'IDxxxxxx.JavaPath' in instance.properties in the directory /usr/sap/<SID>/JCxx/j2ee/cluster
      JProbe relies on the property to find Java home, but there are Web AS 6.30/6.40 SPs that do not have this property.
      If it is not included, add it for dispatcher or server (dependent on which one should be profiled)
      For example, add "ID8156650.JavaPath=C:/java/jdk1.4.2_06."
      Be aware that using the standard startup procedure instance.properties is overwritten, so you might repeat this step.
Step 6: Start the J2EE Engine with JProbe
      Start the database and the central instance (enqueue and message server).
      In JProbe click "Run" for the J2EE settings defined before.
      Dispatcher and server of the J2EE Engine will be started; the one specified in the Integration will be profiled.
SAP Web AS (Java) 6.30/6.40 Start-up: Using JProbe 5.2.x via JLaunch
Starting the Server via JLaunch
Detailed instructions how to start JLaunch can be found in the documentation of SAP Web AS (Java). Here we give only the minimal list of steps:
   1.
      Include the location of JLaunch in the PATH
   2.
      Go to base directory of dispatcher or server (depending on which one you want to start).
      jlaunch pf=c:\usr\sap\<SID>\SYS\profile\<SID>_JCxx_<Host>                   
      -nodename=IDxxxxxxx 
      where "IDxxxxxxx" can be figured out from the "instance.properties" in directory "cluster."
Creating a JProbe Startup File
In order to start up JProbe via JLaunch, you need to create a start-up file (.jpl) with the JProbe launchpad as follows:
   1.
      Click Session -> New J2SE Settings
   2.
      Enter filters, triggers, ...
   3.
      In the Configuration main class and the classpath fields are mandatory; a dummy entry for main class is sufficient, for classpath you could use %CLASSPATH%
   4.
      Save the .jpl file
   5.
      Switch off the Connection Manager; in JProbe 5.2.x a Connection Manager is introduced, which does not work if a Java application is started from a C-framework; therefore use Tools -> Options -> Session -> Advanced to switch off the Connection Manager
   6.
      Change the .jpl file with a text editor
            Comment out the line starting with -jp_java
            Add the connection information and the snapshot directory information; related to the introduction of the Connection Manager this is not written when the .jpl file is saved in the launchpad
                o
                  -jp_socket=<host for analysis>:<port> (e.g. localhost:4444)
                o
                  -jp_snapshot_dir=<snapshot directory> (e.g. c:
Jprobe
snapshots)
Starting JProbe with the Server
Start JLaunch as above, but attach the following parameters to the JLaunch command line:
jlaunch pf=... -nodename=... -Xbootclasspath/a:JProbe-Base-Dir\lib\jpagent.jar
            -Xrunjprobeagent:-jp_input=.jpl-file
For the .jpl-file please, specify the complete path. If there are blanks in the file or directory names, use double quotes or the DOS notation (e.g. PROGRA~1). After starting JLaunch you can now attach the viewer to it (JProbe -> Program -> Attach to Remote/Running Session).
regards
chandra