Looking at 4 recent locklosses, I can't see the reason we lost lock. I investigated using the command line 'lockloss select' and web lockloss tool.

Ryan's 05/01 loclkoss was during work hours (i.e. people around) there is a couple of wobbles in H1:LSC-DARM_IN1 before the lockloss: leftmost plot in attached. I can't see anything similar in the others.

You can see in the attached plot with the IMC that the IMC looses lock ~200ms after AS_A, this means these locklosses are not the "TOGETHER" type we saw (I think exclusively) in O3b: G2201762. 

05/01 17:59UTC 1366999182 Ryan's Monday lockloss.

05/02 09:41UTC 1367055714 After my shift last night

05/02 14:34UTC 1367073283 Before Maintenance began

05/03 00:15UTC 1367108148 This evening