I called Rolf (@LLO) who called Alex (@MIT) who suggested that we power cycle every crate
with an RFM connection as we did before (twice in the past year).
Rob and I followed Yoichi around the lab as he turned off and on everything. There
was no special order; he started at the Y-end and worked his way into the corner and
finishing at the X-End. Along the way we also reset the 2 RFM switches around fb0.
This cured the EPICS problem; the FEs could now boot and received the EPICS data.
However, there are still some residual channel hopping-ish issues which Rob and Yoichi are
now working on. |