Troubleshooting

 

This section contains information that will help you troubleshoot your BEA eLink Adapter into WebLogic Enterprise CORBA configuration. These tips are listed in the order it is suggested that you follow for troubleshooting, although it is not required that you follow this order.

The following topics are discussed:

• General Rules

• Checking Environment Variables

• Booting WLE

• Booting the CORBA Adapter with Maximum Tracing

• Verifying That All Expected Servers and Services are Running on All Platforms

• Verifying That the Client Program Runs

• Verifying That the Configuration Tools Run

 

---

General Rules

The following general rules will help you troubleshoot your system:

• With any error, always check the ULOG file. The ULOG file is a text file that contains error, warning, and informational messages generated by the eLink Platform and components like the eLink Adapter for CORBA.

  The ULOG file name defaults to $APPDIR/ULOG.mmddyy. All but the extension can be changed by setting the ULOGPFX value in the MACHINES section of the UBB Configuration file.

  For example, given the entry:

  ULOGPFX="/work/logs/zlog"

  The log file name on 4 July, 2000, would be:

  /work/logs/zlog.070400

• While running WLE, use the following tmadmin commands:

  • psr - lists active servers.

  • psc - lists active services.

  • q - quits tmadmin.

 

---

Checking Environment Variables

Make sure the environment variables are set correctly by running the WLE buildserver and tmadmin command line utilities.

Try entering the following command lines:

• buildserver

  • If you receive a "not found" error, check the following:
    The TUXDIR environment variable is set and points to the WLE installation directory.

    The PATH environment variable includes $TUXDIR/bin.

    If you receive a "Can't open shared library" error (Unix only), check that the SHLIB_PATH (HP) or LD_LIBRARY_PATH (Solaris) includes $TUXDIR/lib.

    If you receive a message indicating an unlicensed version, make sure that you have a valid lic.txt file in $TUXDIR/udataobj.

• tmadmin

  • If you receive a "TUXCONFIG environment variable not set" error, check that APPDIR and TUXCONFIG environment variables are properly set and TUXCONFIG contains a complete pathname. The usual setting for TUXCONFIG is $APPDIR/tuxconfig.

  • Other errors may indicate that TUXCONFIG points to a non-existing or invalid tuxconfig file.

  • If tmadmin loads, but you get a "No bulletin board exists." message, WLE is not active, you can access the utilities and enter q (quit) to exit tmadmin.

 

---

Booting WLE

If the command line utilities work, most of the global environment variables are set correctly. The next step is to attempt to boot WLE.

• In theMACHINES section of the WLE UBB configuration file, check that the hostname is correct and that the values listed for TUXDIR, TUXCONFIG, and APPDIR are correct and match the values for the global environment variables. Also check that all required servers have valid entries in the SERVERS section.

  Required servers are:

  Table 5-1

  WLE	TMSYSEVT, TMFFNAME, ISL (note that the TMFFNAME server is listed three times, with different CLOPT entries, all three are required)
  NameService	cns (CORBA NameService)
  e2a	TMIFRSVR, e2a
  Sample	simple_server
  Remote Access	DMADM, GWADM, GWTDOMAIN

  Note: Detailed information on UBB Configuration files can be found in the ubbconfig(5) reference pages in the BEA Tuxedo Reference Manual (http://e-docs.beasys.com/wle/tuxedo/refman/sect5/index.htm). Refer also to the sample UBB files that are distributed with WLE.

• Try entering the following command line:

  tmloadcf UBBFileName

  • If you receive a 'tmloadcf cannot run on a non-master node' error, check to see that the host name is correct.

  • If you receive a TUXCONFIG error, make sure the value for TUXCONFIG specified in the MACHINES section of the UBB file agrees with the global version.

  • If you receive 'file open errors' (for example, on attempting to open a ULOG file) make sure that APPDIR and TUXDIR are set correctly.

 

---

Booting the CORBA Adapter with Maximum Tracing

The next step is to boot the system with maximum tracing enabled by setting the MINMSGLEVEL and MAXMSGLEVEL values to 0 and 9 respectively in the Adapter configuration file (this is the file name specified in the CLOPT parameter for e2a).

Enter the following command line:

tmboot

• If nothing boots (not even the BBL), make sure that the TUXDIR and APPDIR are correct.

• If you get 'Cannot exec, executable file not found' error(s), make sure that the server names are correct and can be found either in the current working directory or in the PATH.

• If you get 'Application initialization failure' error(s), check the following entries in the ULOG file:

  • The CLOPT line for the server(s) that failed is correctly formed.

  • Any files specified in the CLOPT line actually exist.

• A message 'Process id=xxxx Assume started (pipe)' can indicate a problem with data, such as a METHOD value in the Adapter configuration file. Check both psr and psc commands in tmadmin and verify that all servers and services are running.

 

---

Verifying That All Expected Servers and Services are Running on All Platforms

If tmboot seemed to work, use the tmadmin psr command to verify that all servers listed in the UBB Configuration file SERVERS section are running (note that psr will not list DEFAULT). Using the tmadmin psc command, also verify that all services listed in the Adapter Configuration file(s) SERVICE section(s) are running. Refer to the ULOG file for possible information about any missing servers or services.

On Windows NT, tmadmin should show the DMADM, GWADM, and GWTDOMAIN servers are running. The required services should show up with a program name of GWTDOMAIN. The following tmadmin output lists services and servers running after booting WLE with e2aremote.ubb.

Listing 5-2 Output from tmadmin for Windows NT

---

> psr
Prog Name      Queue Name  Grp Name      ID RqDone Load Done Current Service
---------      ----------  --------      -- ------ --------- ---------------
BBL.exe        55432       SITE1          0      0         0 (  IDLE )
DMADM.exe      00001.00001 GW_GRP         1      6       300 (  IDLE )
GWADM.exe      00001.00002 GW_GRP         2      0         0 (  IDLE )
GWTDOMAIN.exe  00001.00003 GW_GRP         3      0         0 (  IDLE )

> psc
Service Name Routine Name Prog Name  Grp Name  ID    Machine  # Done Status
------------ ------------ ---------  --------  --    -------  ------ ------
DMADMIN      DMADMIN      DMADM.exe  GW_GRP     1      SITE1       0 AVAIL
LCL_NT       GWS          GWADM.exe  GW_GRP     2      SITE1       0 AVAIL
GWA          GWA          GWADM.exe  GW_GRP     2      SITE1       0 AVAIL
Simpleto_up+ GWS          GWTDOMAIN+ GW_GRP     3      SITE1       0 AVAIL
Simpleto_lo+ GWS          GWTDOMAIN+ GW_GRP     3      SITE1       0 AVAIL
> q

---

On the Unix platform, tmadmin whould show that all servers, including DMADM, GWADM, and GWTDOMAIN are running and that all expected services are available. The following tmadmin output lists services and servers running after booting WLE with e2alocal.ubb.

Listing 5-3 Output from tmadmin for Unix

---

> psr
Prog Name      Queue Name  Grp Name      ID RqDone Load Done Current Service
---------      ----------  --------      -- ------ --------- ---------------
BBL            55432       SITE2          0     17       850 (  IDLE )
DMADM          00003.00001 GW_GRP         1      6       300 (  IDLE )
simple_server  00002.00001 APP_GRP        1      0         0 (  IDLE )
TMSYSEVT       00001.00001 SYS_GRP        1     20      1000 (  IDLE )
GWADM          00003.00002 GW_GRP         2      0         0 (  IDLE )
TMFFNAME       00001.00002 SYS_GRP        2      3       150 (  IDLE )
GWTDOMAIN      00003.00003 GW_GRP         3      0         0 (  IDLE )
TMFFNAME       00001.00003 SYS_GRP        3      1        50 (  IDLE )
TMFFNAME       00001.00004 SYS_GRP        4      0         0 (  IDLE )
TMIFRSVR       00001.00005 SYS_GRP        5     32      1600 (  IDLE )
cns            00001.00006 SYS_GRP        6      0         0 (  IDLE )
e2a            00002.00010 APP_GRP       10      0         0 (  IDLE )

> psc
Service Name Routine Name Prog Name  Grp Name  ID    Machine  # Done Status
------------ ------------ ---------  --------  --    -------  ------ ------
DMADMIN      DMADMIN      DMADM      GW_GRP     1      SITE2       0 AVAIL
REM_UX       GWS          GWADM      GW_GRP     2      SITE2       0 AVAIL
Simpleto_up+ INVOKE_METH+ e2a        APP_G+    10      SITE2       0 AVAIL
Simpleto_lo+ INVOKE_METH+ e2a        APP_G+    10      SITE2       0 AVAIL
>q

---

 

---

Verifying the Remote Connection

After ensuring that all servers and services are running on both platforms, try running the client program on Windows NT:

e2aclient upper "for all good men"

• If you receive an error message, "tpinit() failed - TPESYSTEM - internal system error", WLE (or Tuxedo) is not running on Windows NT.

• If you receive an error message, "Unable to request Service xxx, rc = 10 (TPESVCERR - server error while handling request), check the following:

1. If there is NOT a ULOG message on the Unix platform indicating a service request, check:

   1. The server is actually running on the Unix platform.

   2. The NWADDR in the NT DMCONFIG file for the server is correct.

2. If there is a ULOG message on the Unix platform with an error message "ERROR: Unable to obtain remote domain id (xxx) information from shared memory", or a similar message in the NT ULOG file, check to see that the DOMAINID fields in the DMCONFIG files match.

   NT DMCONFIG                     Unix DMCONFIG
   
   *DM_LOCAL_DOMAINS               *DM_LOCAL_DOMAINS
   LCL_NT DOMAINID="LOCAL_NT"      REM_UX  DOMAINID="REMOTE_UX"
   
   *DM_REMOTE_DOMAINS              *DM_REMOTE_DOMAINS
   REM_UX DOMAINID="REMOTE_UX"     LCL_NT  DOMAINID="LOCAL_NT"

 

---

Verifying That the Client Program Runs

The last troubleshooting step is to see if the client program will run and talk to the server program through the eLink for CORBA Adapter.

Using the sample e2a client program, enter a sample command, such as:

e2aclient upper "now is the time"

• If you receive an error message, "Unable to request Service xxx , rc = 6 (TPENOENT - no entry found)", make sure that you are requesting an advertised service by verifying that the service (see the SERVICENAME parameters in the Adapter configuration file) is actually being advertised (see the ULOG file and use the tmadmin psc command) and match the service alias you are using in the client program.

• If you receive an error message, "Unable to request Service xxx, rc = 11 (TPESVCFAIL - application level)", verify that the FIELDTBLS32 and FLDTBLDIR32 environment variables are correctly defined.

  Note: You must enter a tmshutdown and tmboot sequence for any change to these variables to take effect.

 

---

Verifying That the Configuration Tools Run

The eLink Adapter for CORBA includes two command line tools that are useful in configuring the Adapter. The CFGEN tool generates a skeleton configuration file using the Interface Repository.

The FTGEN tool generates an FML Field Table File using the configuration file and the Interface repository as input. The following troubleshooting examples apply to either the FTGEN or CFGEN utilities, CFGEN is used as an example.

Start the command line tool by typing CFGEN.

• If you get the message 'CFGEN: not found' on the console, verify that the directory where the CFGEN tool is installed (typically $TUXDIR/bin) is in the system PATH.

• If you get the message 'CFGEN[3]: java: not found' on the console verify that the java 1.2 JDK/bin or JRE/bin directory is in the system PATH.

• If you get the message 'Exception in thread "main" java.lang.NoClassDefFoundError: CFGEN on the console, verify that the TUXDIR environment variable is set and points to the directory where WLE is installed.

• If you get any other messages issued by the tools, refer to Error and Information Messages.