[Dave, E.J., Erik]
At about 16:25 UTC, the corner station had a timing glitch that put the IOPs on h1susb123, h1sush2a, h1sush34, h1sush56 and h1lsc0 into an error state.
The affected front ends had no error messages in the system logs..
The SUS IOPs all had long cycle times in the tens of milliseconds. These times are typical of a "dolphin" timing glitch, seen when someone powers on a computer attached to the dolphin network, but no system was powered on in this case.
h1lsc0 had a long cycle of only 3 milliseconds, too short for a "dolphin" glitch. Also, it's time_0 value, which reads the clock time at the start of each second was off by 17 milliseconds, though the duotone value was correct. (see attached screen shot).
There's no evidence anyone was in the CER or MSR at the time of the glitch.
The operator put the IFO into a safe state. Models on the affected systems were restarted.
During the restart, h1susb123 reset it's Ethernet adapter.