Earlier tonight before the ALS problems started, I made a couple of changes to the guardians. 

1) I set the AS36 offsets to 0 when they are turned off before ENGAGE_ASC for full IFO.  This will stop the ASC SDF diffs that operators have been accepting each lock. 

2) I made some changes to the PRMI path. 

• I reduced the time that it tries to lock DRMI from 10 minutes to 5 minutes before the guardian redirects to try locking PRMI and running the ASC. 
• Watching the guardian run through this step as Jeff B was relocking, it locked PRMI then calculated the shortest path to low noise was to go right back to DRMI without running the ASC, so I rearranged some paths. 
• With help from TJ and Jamie I added wieghts of 5 to several paths in this part of the graph, which makes those paths longer in the shortest path calculation.  I think this should now run through the PRMI ASC, offload it, and then move onto DRMI locking. I didn't stay long enough to see this work though.