Relocking after and earthquake last night our range dropped from around 115 to around 107Mpc, this was because a bug in the squeezer LO guardian would under some circumstances not engage the boost in the LO servo . 

TJ noted the range drop and Cheryl looked at some alignment changes: 

• https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=49905
• https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=49912
• Patrick has started some BRUCO scans to compare before and after lock stretches:
  • https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=49913

The problem was that there was a check on the power in the 3MHz beatnote that was executed twice, in a way that left us open to unintended configurations.  This wrong configuration must have also been accepted in SDF. There was probably confusion because there was also a beckhoff crash from which things were not correctly restored last night. 

I've edited the SQZ_LO_LR guardian to avoid this situation in the future.  I also simplified the graph for the LO guardian because there were unneeded and potentially problematic connections in it. (screenshot of the changes to the graph attached). 

The last attachment shows the change in the DARM spectrum when the LO boosts were off, there is excess noise around 100Hz.