My cluster is not receiving a lot of users, but there is a lot of traffic on the wires anyway. The beat period, arp period, or broadcast period for your levelers may be set too high. Experiment with reducing this value in your configuration files and see if that cuts down on the load. The values the tacoplex.conf that comes with the distribution are a good tradeoff between performance and robustness.
I get an "unable to send command to kernel" message whenver I try to do anything. You may be trying to run TacoPleX utilities under a non-TacoPleX kernel. Check to see if there is a file called /proc/net/ip_hr/summary. If not, you need to reboot into the TacoPleX kernel.
Traffic is getting to the levelers but not to the servers. You may have forgotten to enable IP forwarding in the kernel. See the software installation instructions above for how to do this.
I get an error message every time I run tacoplexadm. This is almost certainly a configuration error. Check out the values in your configuration file and make sure that jive with the machine setup. Network problems are the biggest culprits.
I ran tacoplexd but when I checked the process list, it wasn't running. There was an error staring up the daemon. Check your system log file for messages from tacoplexd. If it died in an orderly fashion, there will be a message telling you why it couldn't start. Most startup problems are caused by configuration errors.
I keep getting a page full of register dumps and my machine locks. This is bad news. It means the kernel is crashing. Please send a report of this behavior to the author. Include as much detail as possible about your system configuration and what your cluster doing when the problem occured.