Job Management Partner 1/Software Distribution Client Description and User's Guide

12.2.2 Failures in the distributor tools

Phenomena Method of investigation Cause and corrective actions

The distributor tools cannot start. Check whether an insufficient swap message is displayed on the console. If the message is output, the server process is unable to start because the disk is full.
Change the size of the swap file.

Check whether the KDDI4001-E message is output in the following message log file.

For systems other than HP-UX
/NETMAF/dmprm/MAIN.LOGF

For HP-UX
/var/opt/NETMAF/dmprm/MAIN.LOGF

If the message shown on the left is displayed, this indicates a setting error of the JP1/Software Distribution Client service port.
If the JP1/Software Distribution Client maintenance information 2 shows gaiSKLGetServByName, the port number of the service name netmdm has not been set in the /etc/services file.
For details on how to set the port, see 6.3.1 Setting the port numbers in PART 2.

Cannot connect to the managing server during packaging Check the output messages in the following message log files.

For systems other than HP-UX
/NETMAF/dmprm/MAIN.LOGF

For HP-UX
/var/opt/NETMAF/dmprm/MAIN.LOGF

If the KDDI4012-E message is output:
A connection is rejected due to the flow restriction of JP1/Software Distribution Client.

If a message indicating a server failure is output:
A JP1/Software Distribution Client server failure occurred. In this case, JP1/Software Distribution Client comes to an emergency stop, and becomes inactive.

If the KDDI1009-E or KDDI1013-E message is output:
A TCP/IP network setting error occurred at the managing server. Refer to /etc/hosts to check the following:

Whether the host name of the packager is defined.

Whether the definition of the IP address matches the packager.

For details on how to set the host name, see 6.3.3 Setting the host name in PART 2.

If none of the above message is output:
Use the ping command to check the communication condition of the managing server and the packager. The following shows the command execution results and the cause of the failure.

If the message indicating that the network cannot be reached is output, there may be a setting error in the routing information.

If there is no response to the command, there may be a failure in the network equipment.

If there is a response to the command, there may be a temporary communication failure on the network.
In this case, packaging may be successful if it is retried.

Enter the rdsafse command to start JP1/Software Distribution Client on the relay system, and then check the message that is output.

If the KDDI4013-I message is output:
JP1/Software Distribution Client may not be inactive.

If the KDDI3064-W message is output:
JP1/Software Distribution Client is running. Other causes may be the problem.

A communication error (file transfer error) occurs during packaging Check the following message log files.

For systems other than HP-UX
/NETMAF/dmprm/MAIN.LOGF

For HP-UX
/var/opt/NETMAF/dmprm/MAIN.LOGF

If a message that indicates an occurrence of a server failure is output, a server failure occurred in JP1/Software Distribution Client. In this case, JP1/Software Distribution Client comes to an emergency stop, and becomes inactive.

For the /etc/passwd file of the managing server and packager, use the ls command to check the following.

Whether the netmdms user is registered.

Whether the user ID and the ID group of the netmdms user matches between systems.

If the items on the left are not set properly, the user ID is incorrectly registered.
Reset and retry.
For details on how to set user ID, see 6.3.2 Setting the user ID in PART 2.

Use the cat command to check the contents of the following login permission file, and make sure that the local host name is registered.

For systems other than HP-UX
/NETMRDS/buff/.rhosts

For HP-UX
/etc/opt/NETMDMW/buff/.rhosts

If the local host name is not registered, re-register it.

Execute the ping command. If the ping command returns a response properly, there may be a temporary communication failure on the TCP/IP network.
In this case, packaging may be successful if it is retried.

During distribution from JP1/Software Distribution Manager, the job remains in the WAIT (waiting for execution) status. Use the rdssts command of the client services to check whether the package distribution command is being executed from the client services. If the package distribution command is being executed, the distribution execution request may have been refused because the client services were distributing packages to another managing server.
Execute the rdsafind -x command after the command is completed.

For the /etc/passwd file of the managing server and distribution-destination system, use the ls command to check the following:

Whether the netmdms user is registered.

Whether the user ID and the ID group of the netmdms user matches between systems.

If the items on the left are not set properly, the user ID is incorrectly registered.
Reset and retry.
For details on how to set the user ID, see 6.3.2 Setting the user ID in PART 2.

Use the cat command to check the contents of the following login permission file, and make sure that the local host name is registered.

For systems other than HP-UX
/NETMRDS/buff/.rhosts

For HP-UX
/etc/opt/NETMDMW/buff/.rhosts

If the local host name is not registered, re-register it.

During distribution from JP1/Software Distribution Manager, the job status remains in the EXEC (executing) status. Check the operation status of the distribution-destination system. The power of the distribution-destination system may be off.

Check the following setting backup file to see whether the appropriate monitoring time is set for WatchTimeofStatus in the settings file of the client services.

For systems other than HP-UX
/NETMRDS/rdsprm/RDSSETFL.bak

For HP-UX
/etc/opt/NETMDMW/rdsprm/RDSSETFL.bak

Correct the monitoring time as necessary.

Note

The relay system monitors the result of distribution to the distribution-destination systems beneath itself at regular intervals, and reports this to the managing server. If the monitoring time is too long, the system may remain in the executing status even when the distribution is completed.

For example, if the default is set for WatchTimeofStatus, monitoring is only performed once every 60 minutes. In this case, after distribution is complete, the system may remain in the executing status for up to 60 minutes.

If you failed to start lower-level systems, and then restarted it, check that the specification method of the distribution destination (host name or IP address specification) is not changed. If the specification method of the distribution destination has been changed, use the rdsdmind command of JP1/Software Distribution Manager to delete or execute the remaining job.

During distribution from JP1/Software Distribution Manager, the job terminates in an error. Use the rdsafind command (with the -l option), view the notified information from the distribution-destination system. The probable causes are:

Communication error with the distribution-destination system.

Improper installation conditions (disk is full) at the distribution destination system.

For details on the status codes, see the manuals for individual distribution-destination systems.

Phenomena	Method of investigation	Cause and corrective actions
The distributor tools cannot start.	Check whether an insufficient swap message is displayed on the console.	If the message is output, the server process is unable to start because the disk is full. Change the size of the swap file.
Check whether the KDDI4001-E message is output in the following message log file. For systems other than HP-UX /NETMAF/dmprm/MAIN.LOGF For HP-UX /var/opt/NETMAF/dmprm/MAIN.LOGF	If the message shown on the left is displayed, this indicates a setting error of the JP1/Software Distribution Client service port. If the JP1/Software Distribution Client maintenance information 2 shows gaiSKLGetServByName, the port number of the service name netmdm has not been set in the /etc/services file. For details on how to set the port, see 6.3.1 Setting the port numbers in PART 2.
Cannot connect to the managing server during packaging	Check the output messages in the following message log files. For systems other than HP-UX /NETMAF/dmprm/MAIN.LOGF For HP-UX /var/opt/NETMAF/dmprm/MAIN.LOGF	If the KDDI4012-E message is output: A connection is rejected due to the flow restriction of JP1/Software Distribution Client.
If a message indicating a server failure is output: A JP1/Software Distribution Client server failure occurred. In this case, JP1/Software Distribution Client comes to an emergency stop, and becomes inactive.
If the KDDI1009-E or KDDI1013-E message is output: A TCP/IP network setting error occurred at the managing server. Refer to /etc/hosts to check the following: Whether the host name of the packager is defined. Whether the definition of the IP address matches the packager. For details on how to set the host name, see 6.3.3 Setting the host name in PART 2.
If none of the above message is output: Use the ping command to check the communication condition of the managing server and the packager. The following shows the command execution results and the cause of the failure. If the message indicating that the network cannot be reached is output, there may be a setting error in the routing information. If there is no response to the command, there may be a failure in the network equipment. If there is a response to the command, there may be a temporary communication failure on the network. In this case, packaging may be successful if it is retried.
Enter the rdsafse command to start JP1/Software Distribution Client on the relay system, and then check the message that is output.	If the KDDI4013-I message is output: JP1/Software Distribution Client may not be inactive. If the KDDI3064-W message is output: JP1/Software Distribution Client is running. Other causes may be the problem.
A communication error (file transfer error) occurs during packaging	Check the following message log files. For systems other than HP-UX /NETMAF/dmprm/MAIN.LOGF For HP-UX /var/opt/NETMAF/dmprm/MAIN.LOGF	If a message that indicates an occurrence of a server failure is output, a server failure occurred in JP1/Software Distribution Client. In this case, JP1/Software Distribution Client comes to an emergency stop, and becomes inactive.
For the /etc/passwd file of the managing server and packager, use the ls command to check the following. Whether the netmdms user is registered. Whether the user ID and the ID group of the netmdms user matches between systems.	If the items on the left are not set properly, the user ID is incorrectly registered. Reset and retry. For details on how to set user ID, see 6.3.2 Setting the user ID in PART 2.
Use the cat command to check the contents of the following login permission file, and make sure that the local host name is registered. For systems other than HP-UX /NETMRDS/buff/.rhosts For HP-UX /etc/opt/NETMDMW/buff/.rhosts	If the local host name is not registered, re-register it.
Execute the ping command.	If the ping command returns a response properly, there may be a temporary communication failure on the TCP/IP network. In this case, packaging may be successful if it is retried.
During distribution from JP1/Software Distribution Manager, the job remains in the WAIT (waiting for execution) status.	Use the rdssts command of the client services to check whether the package distribution command is being executed from the client services.	If the package distribution command is being executed, the distribution execution request may have been refused because the client services were distributing packages to another managing server. Execute the rdsafind -x command after the command is completed.
For the /etc/passwd file of the managing server and distribution-destination system, use the ls command to check the following: Whether the netmdms user is registered. Whether the user ID and the ID group of the netmdms user matches between systems.	If the items on the left are not set properly, the user ID is incorrectly registered. Reset and retry. For details on how to set the user ID, see 6.3.2 Setting the user ID in PART 2.
Use the cat command to check the contents of the following login permission file, and make sure that the local host name is registered. For systems other than HP-UX /NETMRDS/buff/.rhosts For HP-UX /etc/opt/NETMDMW/buff/.rhosts	If the local host name is not registered, re-register it.
During distribution from JP1/Software Distribution Manager, the job status remains in the EXEC (executing) status.	Check the operation status of the distribution-destination system.	The power of the distribution-destination system may be off.
Check the following setting backup file to see whether the appropriate monitoring time is set for WatchTimeofStatus in the settings file of the client services. For systems other than HP-UX /NETMRDS/rdsprm/RDSSETFL.bak For HP-UX /etc/opt/NETMDMW/rdsprm/RDSSETFL.bak	Correct the monitoring time as necessary. Note The relay system monitors the result of distribution to the distribution-destination systems beneath itself at regular intervals, and reports this to the managing server. If the monitoring time is too long, the system may remain in the executing status even when the distribution is completed. For example, if the default is set for WatchTimeofStatus, monitoring is only performed once every 60 minutes. In this case, after distribution is complete, the system may remain in the executing status for up to 60 minutes.
If you failed to start lower-level systems, and then restarted it, check that the specification method of the distribution destination (host name or IP address specification) is not changed.	If the specification method of the distribution destination has been changed, use the rdsdmind command of JP1/Software Distribution Manager to delete or execute the remaining job.
During distribution from JP1/Software Distribution Manager, the job terminates in an error.	Use the rdsafind command (with the -l option), view the notified information from the distribution-destination system.	The probable causes are: Communication error with the distribution-destination system. Improper installation conditions (disk is full) at the distribution destination system. For details on the status codes, see the manuals for individual distribution-destination systems.

[Contents][Back][Next]

[Trademarks]