Quoting jaharkes_at_cs.cmu.edu (jaharkes_at_cs.cmu.edu):
> now, the only `stable' solution is to either do everything connected, or
> have a very short reintegration interval. ('cfs wd -age 5 -time 5')

This seems to have gotten me past the first hurdle. Next, what about these,
are they something to be concerned about, or are they purely informational?
(And is the a TFM to pull these answers out of, lest I waste peoples' time?)

[21:13:45]SocketListener: "sl.c", line 435:    DecodePacket(RPC2_REPLY): state != AWAIT
[21:13:52]SocketListener: "sl.c", line 435:    DecodePacket(RPC2_REPLY): state != AWAIT
[21:14:12]SocketListener: "sl.c", line 435:    DecodePacket(RPC2_REPLY): state != AWAIT
[21:14:25]SocketListener: "sl.c", line 439:    DecodePacket(RPC2_REPLY): bad seqno
[21:15:27]SocketListener: "sl.c", line 435:    DecodePacket(RPC2_REPLY): state != AWAIT
[21:15:31]SocketListener: "sl.c", line 435:    DecodePacket(RPC2_REPLY): state != AWAIT
[21:16:05]SocketListener: "sl.c", line 435:    DecodePacket(RPC2_REPLY): state != AWAIT
[21:16:10]SocketListener: "sl.c", line 435:    DecodePacket(RPC2_REPLY): state != AWAIT

Mike Stone