We noticed that H1ISIBS had gone down and check the dmesg logs to find this: 
[29807.260951] h1iopseib2: WARN - Missed an expected DAC sample, DAC card: 1, chan 13, time_sec: 1469407411, cycle: 32769
[29808.620352] h1isibs: ERROR - Waiting for ADC cycle 32737, live read is 53217, mm: 0, ioMemCtr: 4065
[29808.620529] h1isibs: ERROR - An ADC timeout error has been detected, waiting for an exit signal.
[31724.338050] perf: interrupt took too long (3129 > 3127), lowering kernel.perf_event_max_sample_rate to 63750 

Jonathan restarted the model ~ around 2:15 UTC -ish.