Recurring IPC error on h1lsc0 suggests problem with IPC timing
h1susprocpi gets a recurring error almost always one per second from h1omc, running on heavily loaded and fast-running h1lsc0.
The steady nature of the error suggests that it's not random, which further suggests that there may be a particular point in a second where the error is likely to occur, at the second boundary perhaps.
Shutting down an unrelated model on h1lsc0 lessons the error rate, but does not totally eliminate it.
h1omc, the sender, is a 16k system. h1susprocpi, the receiver, is a 2k system. Both are finishing their cycles in plenty of time. Since IPCx senders always write output to the next cycle, this should never fail unless the sender has a long cycle. h1omc does not.