HPlogo   Configuring and Managing MPE/iX Internet Services:
HP 3000 MPE/iX Computer Systems > Chapter 2 Internet Daemon

2.6: Troubleshooting inetd

MPE documents

Complete PDF

 

Table of Contents

Glossary

Index

 

⇓ Page Bottom

⇑ Page Top

  

Using inetd Message Logging

Inetd Implementation Differences

This section explains the kinds of error messages you may see regarding the operation of inetd. The messages will appear either on the console or they will be sent to the $STDLIST for inetd or both, depending upon the message's level of importance.

Message Explanation
An inetd is already running You attempted to start inetd when one is already running. You may invoke inetd a second time if you use the -c, -k, or -l option, but you cannot run multiple copies of inetd.
There is no inetd running You attempted to reconfigure inetd when none was running. The first time you run inetd, you must stream it as a job or run it interactively without specifying the -c (reconfiguration) option.
Inetd not found This message occurs if you invoke inetd with the -c option and inetd cannot reread its configuration file (which is the purpose of -c). This occurs when the original Internet daemon dies or is killed without releasing its semaphore. (The Internet daemon locks a global semaphore to indicate when it is running to prevent users from running more than one inetd at a time.) To fix the problem, enter the inetd -k command to remove the semaphore left by the previous Internet daemon, then restart inetd.

The following diagnostic messages are generated by successful and failed attempts to establish a connection to the Internet Services.

Message Explanation
/etc/inetd.cnf: Unusable configuration file

The Internet daemon cannot access its configuration. The error message preceding this one specifies the reason for the failure.

/etc/inetd.conf: line number: nnn error There is an error on the line specified by nnn in the inetd configuration file. The Internet daemon skips this line, continues reading the rest of the file, and configures itself accordingly. To solve the problem, open the configuration file, edit the erroneous line, and save the corrected version. Then, tell inetd to reread the new version of INTEDCNF by issuing the inetd.net.sys -c command at the CI prompt.
system call: The system call noted in the error message failed. See the corresponding entry in the Berkeley Sockets/iX Reference Manual for a description of the system call. The reason for the failure is explained in the error message appended to the system call name.
Cannot configure inetd Due to errors in the inetd configuration file, none of the services it lists could be set up properly.
Too many services running The maximum number of services allowed to access inetd simultaneously has been exceeded.
file: found before end of the line An entry in a configuration file may need to exceed one line. If so, you indicate that the line continues by inserting a backslash at the end, then continue typing data on the next line. If, however, you place a backslash in the middle of the line, inetd will ignore it and continue reading to the end of the current line, but will not continue to the next line. In this case, it is likely that the configuration information will be misread.
service/protocol; Unknown service The system call getservbyname failed because the service is not listed in the services file. To solve the problem, you may either add an entry for the service to the services file or delete the entry for the service from the inetd configuration file.
service/protocol: Server failing (looping), service terminated. When inetd tries to start 40 servers within 60 seconds for a datagram service, it assumes that the server is failing to handle the connection. To avoid entering a potentially infinite loop, inetd issues this message, discards the packet requesting the socket connection, and refuses further connections for this service. After 10 minutes, inetd tries to reinstate the service and accept connection requests.
service/protocol: socket service/protocol: listen service/protocol: getsockname Any of these three errors renders the service unusable. To make the service available again, you must issue the inetd -c command to have inetd reread the configuration file.
service/protocol: bind: Indicates that the service is temporarily unusable because inetd cannot bind the service to the socket. After 10 minutes, inetd tries to bind the socket again. If it is successful, then it will listen for a connection request and provide the appropriate service. If it fails, it will wait another 10 minutes and try again.
service/protocol: Access denied to remote host (address) The remote host failed to pass the security test for the service indicated in the message. If this message appears frequently, it can indicate that someone is trying to repeatedly access your system, and failing.
service/protocol: Connection from remote host (address) When connection logging is enabled, this message indicates a successful connection attempt to the specified service.
service/protocol: Added service, server executable Records the services that are added when you reconfigure inetd.
service/protocol: New Lists the new user identifications, new servers, or executable programs used for the service when reconfiguring inetd.
service/protocol: Deleted service Records the services that are deleted when you reconfigure inetd.

The following diagnostic and error messages are generated by problems in the inetd security file.

Message Explanation
/usr/adm/inetd.sec: Field contains other characters in addition to * for service The wildcard character (*) is used in combination with additional integer(s) in one part of an address field, which is not allowed. For example, the Internet address 10.5*.8.7 entered in the inetd security file will generate an error message because the second field includes a 5 followed by the * character. Either integers or the wildcard character is allowed in one part of an address field.
/usr/adm/inetd.sec: Missing low value in range for service You have used the range indicator (-) in the wrong way in an entry in the inetd security file. For example, the second field of the Internet address 10.-5.8.7 is incorrect because it does not include both a starting range number ("high value") and the ending range number ("low value"). A correct use of the range indicator in an Internet address would be 10.8-5.8.7.
/usr/adm/inetd.sec: Missing high value in range for service You have used the range indicator (-) in the wrong way in an entry in the inetd security file. For example, the second field of the Internet address 10.5-.8.7 is incorrect because it does not include both a starting range number ("high value") and the ending range number ("low value"). A correct use of the range indicator in an Internet address would be 10.8-5.8.7.
/usr/adm/inetd.sec: High value in range is lower than low value for service You expressed a range of numbers incorrectly in an entry in the inetd security file. For example, the second field of the Internet address 10.5-8.8.7 is incorrect because the starting range number ("high value") is lower than the ending range number ("low value"). A correct use of the range indicator in an Internet address would be 10.8-5.8.7.
/usr/adm/inetd.sec: allow/deny field does not have a valid entry for service. The entry in the second column is not one of the keywords allow or deny. The inetd server ignores the entry and does not implement security for this service unless there is a subsequent entry in the inetd security file for this service that is correct.


Using inetd Message Logging

Inetd Implementation Differences