I am not whining about the list, I am actually quite please with how it works. At the risk of getting massively flamed for not knowing why this won't work; this looks like a problem I see in real time systems from time to time. If you have a large ongoing task and a simple infrequent task you can speed up the little task lots and still get good overall throughput. Observation 1) I have picked up that the TCLUG list is a very light load for the host system. Observation 2) The injection task is already slated to take a very long time so a few days added to the overall task time may be acceptable. Conjectural question) In practical terms, is it possible to gate off the kernel injection task using cron from time to time, say every hour until the TCLUG message queue is empty? Mark Browne