[Digital logo]
[HR]

DECnet-Plus
Problem Solving


Previous | Contents

The topics in this chapter are:

4.1 Local IP Address Displays As 0.0.0.0

Some TCP/IP products do not support a "read local address" function through the PWIP (PATHWORKS Internet Protocol) driver interface. As a workaround, OSI transport tells Session Control/Node Agent that the node local IP address is 0.0.0.0.

Therefore, the following address is correct:

NCL> SHOW NODE 0 address 
 
     Address                           = 
       { 
          ( 
          [ DNA_CMIP-MICE ] , 
          [ DNA_SessionControlV3 , number = 19 ] , 
          [ DNA_OSItransportV1 , 'DEC0'H ] , 
          [ DNA_OSInetwork , 49::00-33:AA-00-04-00-FF-FF:21 ] 
          ) , 
          ( 
          [ DNA_CMIP-MICE ] , 
          [ DNA_SessionControlV2 , number = 19 ] , 
          [ DNA_OSItransportV1 , 'DEC0'H ] , 
          [ DNA_IP , 0.0.0.0 ] 
          ) , 
          ( 
          [ DNA_CMIP-MICE ] , 
          [ DNA_SessionControlV3 , number = 19 ] , 
          [ DNA_NSP ] , 
          [ DNA_OSInetwork , 49::00-33:AA-00-04-00-FF-FF:20 ] 
          ) 
       } 

4.2 Troubleshooting

If you have problems getting DECnet over TCP/IP to start up properly, check the following:

  1. Verify that you have an OSI transport template with network service attribute defined as RFC 1006.
    Issue the command:
    NCL> SHOW OSI TRANSPORT TEMPLATE * WITH NETWORK SERVICE = rfc1006 
    

    If you do not have a template defined, then you must execute NET$CONFIGURE Option 4 and replace your OSI transport startup script.
  2. Verify that you have started TCP/IP, and that your product supports the PWIP interface. If you are using Digital TCP/IP Services for OpenVMS, be sure that you have executed the following command procedure:
    SYS$STARTUP:UCX$PWIP_STARTUP.COM 
    
  3. Verify that the PWIP interface is properly registered. Using the management tool of the TCP/IP product installed, verify that the RFC 1006 listener ports defined in OSI transport are known by TCP/IP.
    If you are running Digital TCP/IP Services for OpenVMS, use the following command:
    $ UCX SHOW DEVICE 
      
                                Port                       Remote 
    Device_socket  Type    Local  Remote  Service           Host 
     
      bg3         STREAM      23       0  TALENT           0.0.0.0 
      bg4         DGRAM      520       0                   0.0.0.0 
      bg7         STREAM     399       0                   0.0.0.0 
      bg9         STREAM     102       0                   0.0.0.0 
    

    In this case, we are looking for the two listen ports 399 and 102.
    If IP addresses work and IP names do not, use your TCP/IP management tool to verify that your BIND server knows about the name.


Chapter 5
Solving DECnet-Plus Application Problems

This chapter describes how to isolate and correct common DECnet-Plus application problems.

Definition

In this chapter, the term DECnet-Plus applications refers to FTAM, Virtual Terminal (VT), and any application that uses the OSI Applications Kernel (OSAK) application programming interface (API).

This manual does not discuss other applications considered to be OSI applications, such as X.400 products.

Topics in This Chapter

The topics in this chapter are:

5.1 Underlying Components for DECnet-Plus Applications (OpenVMS Only)

Figure 5-1 shows the underlying DECnet-Plus components that DECnet-Plus applications on OpenVMS systems use. Use this information as a guide during fault isolation.

Figure 5-1 Underlying DECnet-Plus Components (OpenVMS)



5.2 Underlying Components for DECnet/OSI Applications Digital UNIX Only

Figure 5-2 shows the underlying DECnet/OSI components used by DECnet/OSI applications on Digital UNIX systems. Use this information as a guide during fault isolation.

Figure 5-2 Underlying DECnet/OSI Components (Digital UNIX)



5.3 Symptoms of DECnet-Plus Application Problems

The following tables show the symptoms of possible DECnet-Plus application problems on both OpenVMS and Digital UNIX systems, on OpenVMS systems only, and on Digital UNIX systems only.

5.3.1 Problem Symptoms for All Systems

Table 5-1 describes DECnet-Plus application problems that can occur on either OpenVMS or Digital UNIX systems.

Table 5-1 Symptoms and Problems for All Systems
Symptom Possible Problem See:
Connection attempts fail. Reasons for failure include:
  • User name, password, or both used in commands are incorrect.
  • A responder is not listening on the remote address that the initiator specified.
  • Errors exist in the local database file.
Section 5.11 or 5.13.
An FTAM or Virtual Terminal responder fails. Reasons for failure include:
  • The requested transport provider is not available.
  • The command syntax used to start an operation was entered incorrectly.
  • An error exists in the format specified in the command line or the local database file.
  • Multiple OSI applications are using the same transport selector.
Section 5.11, 5.14.1, or 5.16.
An FTAM file does not have the expected attributes. Requested format is not supported and must be converted. Section 5.12.2.
A user-written OSI application terminates unexpectedly. Reasons for failure include:
  • Protocol
  • Network
  • Loss of network connection
  • Remote entry aborted connect
Section 5.19 or FTAM API documentation.
A problem with a user-written application affects the OSAK software. There is a coding error in the application or in the way the application uses the OSAK application programming interface. Section 5.19 or OSAK documentation.

5.3.2 Problem Symptoms for OpenVMS Systems Only

Table 5-2 describes DECnet-Plus application problems that can occur on OpenVMS systems only.

Table 5-2 Symptoms and Problems for OpenVMS
Symptom Possible Problem See:
Connection attempts fail. Reasons for failure include:
  • OpenVMS environment is not set up correctly.
  • FTAM application is not set up correctly.
  • A responder is not listening on the remote address that the initiator specified.
  • A remote entity sent a refuse PDU.
Section 5.13, 5.17, or 5.15.
An FTAM responder fails. There is a problem with the OSAK$SERVER_V3. Section 5.13.
A VT responder terminates unexpectedly. Resources are exhausted. Section 5.15.
A VT responder will not start. Another responder is already running. Section 5.15.

5.3.3 Problem Symptoms for Digital UNIX Systems Only

Table 5-3 describes DECnet/OSI application problems that can occur on Digital UNIX systems only.

Table 5-3 Symptoms and Problems for Digital UNIX
Symptom Possible Problem See:
Connection attempts fail. Reasons for failure include:
  • A listener and responder need to be running on the same system (for example, for a gateway node).
  • A listener/responder is not listening on the remote address that the initiator specified.
  • Syntax for alias entry in the local database file /etc/isoapplications is wrong.
  • An incorrect transport template was specified in the local database file /etc/isoapplications.
  • A remote entity sent a refuse PDU.
Section 5.14.
An FTAM or VT responder fails. There is a problem with the ftam_listener or vt_listener process. Section 5.16.

5.4 Isolating DECnet-Plus Application Faults

Before trying any correction procedures, confirm that you have a DECnet-Plus application problem. Use an application other than the one that failed, such as dlogin or set host, to try to establish a connection to the remote node.

If you can establish a connection, a problem with the failed application is likely the cause; use the trace utilities for further fault isolation. If you cannot establish a connection, the problem is probably not application-specific; use the network reachability tests described in Chapter 3, or check the underlying components that the application uses.

5.4.1 Tools to Use

Table 5-4 shows the tools you can use to isolate DECnet-Plus application faults.

Table 5-4 Tools for Isolating Application Faults
For Problems with: Use: And See:
Any DECnet-Plus application Error messages:
  • Check error messages displayed on the user's terminal.
  • Check error messages displayed on the system console.
  • On Digital UNIX, check the /usr/adm/syslog.dated/*/daemon.log file.
  • On OpenVMS for FTAM, check osif$responder.log file located in your default login directory.
OSI application documentation and Section 5.5.
Tracing utility, ositrace. Section 5.8.
DECnet-Plus applications on OpenVMS systems Event logging. OSAK, FTAM and Virtual Terminal documentation.
FTAM and Virtual Terminal on OpenVMS systems OSAK application database and sys$system:isoapplications.dat file. FTAM/VT documentation.
Applications that use OSAK software OSAK trace utility. OSI application documentation or the OSAK programming documentation.

5.4.2 References

Refer to your FTAM and Virtual Terminal documentation for further information about these applications. Refer to your OSAK documentation for information about applications that utilize the OSAK API.

5.5 Using Event Logging and Log Files

On OpenVMS systems, event logging and log files help you isolate FTAM and Virtual Terminal problems. After you enable logging, the system writes events to the sys$manager:operator.log file.

On Digital UNIX systems, FTAM and Virtual Terminal error messages appear on the operator's console. The system writes responder error messages to the /usr/adm/syslog.dated/*/daemon.log file. You do not need to do anything to enable error logging.

5.5.1 Enabling Network Event Logging (OpenVMS Only)

Enabling network event logging on an OpenVMS system can help you quickly identify FTAM or Virtual Terminal problems. Do the following to enable event logging of network and license events:
Step Action
1 Enable OPER privileges with the following DCL command:

$ set process/priv=oper

2 Ensure that OPCOM is running. If you do not know how to do this, get help from your system manager.
3 Do one of the following:
  • To display subsequent errors on a dedicated terminal, enable error logging to the terminal by entering the following DCL command:

    $ reply/enable=(network,license)

  • Type the log file sys$manager:operator.log and examine it.
4 For events at the Data Link layer, specify the event types you want. For both X.25 and IEEE 802.3 events, see your NCL reference documentation. For X.25 events only, see your OSI Transport and X.25 documentation.

5.5.2 Using the FTAM Responder Log File (OpenVMS Only)

Each FTAM responder process on an OpenVMS system creates a log file of its activity during a process. If you are trying to track the responder's operations, this log file may contain pertinent information, depending on where the problem occurs.

Normally, a responder process creates a log file in the SYS$LOGIN directory of the account specified by the user on the FTAM command line.

Responder Log Set Up

Ensure that your system is set up correctly to use the responder log:
Step Action
1 Check for WORLD:READ protection on osif$responder.com.
2 Check that the responder has WRITE access (for osif$responder.log) into the login account's default directory.
3 If you are using an account other than the default FTAM responder account, make sure that the account has a BYTLM of more than 20,000. If not, you may experience "service provider abort" errors.

5.6 Isolating Faults Using Management Tools (OpenVMS Only)

You can use management tools to isolate FTAM or VT faults on OpenVMS systems. Do the following when you use management tools for this purpose:
For this Connection: Do the Following: To Verify:
Inbound to a given address (FTAM only) Use NCL commands to check inbound address in OSAK application database:

show osak application

The following values:
  • Presentation selector
  • Session selector
  • Transport selector
  • File name of responder command file
  • User name
  • Login password of user name (if any) (you need system privileges to do this)
  • Account name (optional)
  • AP-title (optional, not verified)
  • AE-qualifier (optional, not verified)
Outbound to a given address Examine sys$system:isoapplications.dat The following values:
  • Alias
  • Presentation selector
  • Session selector
  • Transport selector
  • OSI transport address (NSAP, TSAP, Provider)
  • AP-title (optional)
  • AE-qualifier (optional)
  • Session version (optional)
  • All field delimiters

5.7 Tracing Overview

Tracing allows you to examine FTAM and Virtual Terminal connections to other OSI systems. The trace operation traces the protocol data units (PDUs) that the Application, ACSE, Presentation, and Session layers send or receive. Tracing is a two-step process. First, you create a binary trace file during a regular FTAM or VT connection. Then you use the ositrace utility to convert the binary file to a readable text file.

5.7.1 Trace Files

The ositrace utility formats the binary trace information. You can redirect output into a file using appropriate mechanisms. By default, the file is written to the default output device.

The ositrace utility creates a new file for each trace in the default directory of the process where tracing is enabled. It also creates a new version of the trace file for each new connection it traces.

5.7.2 Security Information in Tracing

For either initiator or responder traces, if a file specification contains security information, the output for an FTAM or VT Protocol Control Information (PCI) trace contains the initiator identity (which maps to a user name) and the filestore password (which maps to a login password) in plain ASCII text. Trace files containing this information should be securely stored or deleted immediately after the trace data is analyzed.

5.8 Tracing Outbound FTAM and Virtual Terminal Connections

This section describes how to trace FTAM and Virtual Terminal connections to other OSI initiator systems. The original file that the trace operation creates contains binary information. The ositrace utility converts the binary file to a readable text file.
For OpenVMS Systems: For Digital UNIX Systems:
define OSAK_TRACE ON setenv OSAK_TRACE on

The trace file is created in your current working directory with the name init_xxxx.bin, where xxxx is the time and date.

An alternate method of generating a trace for FTAM is to create the following definitions in the initiator process (OpenVMS only):

define osif$trace_enable FTAM_PCI, ACSE_PCI, PRESENTATION_PCI, SESSION 
define osif$trace osif$init.trace 

This creates the file osif$init.trace. It contains the same information as the trace generated by setting OSAK_TRACE to ON, and you can use ositrace to generate a readable form.

5.8.1 Generating a Readable Trace File

The command syntax for starting the ositrace utility is:
For OpenVMS Systems: For Digital UNIX Systems:
ositrace input-file [output-file] ositrace [options] input-file [output-file]

The input-file specifies the binary trace file that is created when you define the trace logical names on OpenVMS systems or specify the -T option, or define the environment variables, on Digital UNIX systems.

The output-file redirects the output to the specified file instead of displaying the output on the default device or file.

On OpenVMS, define OSITRACE:

$ OSITRACE:== $OSITRACE 

5.8.2 ositrace Command Options (Digital UNIX Only)

The trace records of all the components are processed by default if you do not specify one or more of the command options. Table 5-5 describes the ositrace command options on Digital UNIX systems.

Table 5-5 ositrace Command Options
Option Description
-h Displays a brief help message that includes usage syntax and valid command options.
-F Displays trace records only for the FTAM components. The trace monitors protocol control information (PCI) and file-access data unit (FADU) components.

An FTAM PCI trace monitors the inbound and outbound FTAM service primitives. The trace output logs PCI octets and analyzes the FTAM PCI.

An FTAM FADU trace monitors the inbound and outbound file structure data. The trace is formatted into two columns: the left column logs the octets of data and the right column logs the text equivalents of the octets.

-V Displays trace records only for the Virtual Terminal components. The trace monitors protocol control information (PCI).

A Virtual Terminal PCI trace monitors the inbound and outbound Virtual Terminal service primitives. The trace output logs PCI octets and analyzes the Virtual Terminal PCI.

The trace is formatted into two columns: the right column logs the octets of data and the left column logs the text equivalents of the octets.

-A Displays trace records only for the ACSE layer. The trace monitors the inbound and outbound ACSE service primitives. The trace output logs PCI octets and analyzes the ACSE PCI.
-P Displays trace records only for the Presentation layer. The trace monitors the inbound and outbound presentation service primitives. The trace output logs presentation PDU (PPDU) octets and analyzes the presentation PCI.
-S Displays trace records only for the session layer. The trace monitors the inbound and outbound session service primitives. The trace output logs session PDU (SPDU) octets and analyzes the session PCI.
-f Filters timestamps. This option causes all timestamps to appear as the string xx-xxx-xxxx,xx:xx:xx and allows the comparison of two trace files using the diff command.

5.9 Tracing Inbound FTAM and Virtual Terminal Connections

This section describes how to trace FTAM and Virtual Terminal connections from other OSI systems.

5.9.1 Inbound FTAM and VT Tracing (Digital UNIX)

5.9.2 Inbound FTAM Tracing (OpenVMS)

Edit the file sys$system:osif$responder.com to contain the following line:

$ define osak_trace on 

The file resp_xxxx.bin is created in the default directory of the responder process, where xxxx is the time and date.

Alternatively, you can edit the file osif$responder.com to uncomment the following lines:

$! DEFINE /LOG osif$trace_enable FTAM-PCI,ACSE-PCI,PRESENTATION-PCI,SESSION 
$! DEFINE /LOG osif$trace osif$responder.trace 

The next inbound connection causes an FTAM trace file called osif$responder.trace to be created in the default login directory (SYS$LOGIN) of the specified account.

To trace the DAP gateway, you can choose one of the following options:

  1. Define osak_trace in the file sys$system:osif$gtwy_login.com.
  2. Define osif$trace and osif$trace_enable in the file sys$system:osif$gtwy_login.com.


Previous | Next | Contents | [Home] | [Comments] | [Ordering info] | [Help]

[HR]

  PS_PROFILE_004.HTML
  OSSG Documentation
   2-DEC-1996 12:34:17.96

Copyright © Digital Equipment Corporation 1996. All Rights Reserved.

Legal