Tracing suspect problems in routed

If you suspect problems with the routing daemon, routed(1Mtcp), you can log its actions and packet transfers.

To create a trace file for routed:

1. Specify a trace file by adding the -T option and its pathname to the entry in /etc/inet/config that starts routed:

   sort-key:/usr/sbin/in.routed:/usr/sbin/in.gated:y::options -T tracefile
   

   Note that you should preserve the existing routed options.

2. Specifying the -T option increases routed's debug level to 1. This causes it to log only its actions.

   To trace more information, specify one or more -t options in addition to the -T option:

   -t -T tracefile

   Trace actions and packets.

   -t -t -T tracefile

   Trace actions and history of packets and contents after change.

   -t -t -t -T tracefile

   Trace all actions, packets and contents.

3. Find the process ID (PID) of the routing daemon:
   
   ps -ef | grep in.routed

4. Kill the routing daemon with the PID (pid) found from the previous step:
   
   kill pid

5. Shut down and reboot the system.

As before, pid is the process ID of routed found using ps.

To decrease the trace logging level, send the SIGUSR2 signal to routed:

kill -USR2 pid

You can also use the rtquery command to change the trace logging level of routed and obtain a dump of the daemon's routing table. See the rtquery(1Mtcp) manual page for more information.