Attached are two 270-day trends of the HPO diode box powers (in relative %, first attachment) and the 35W FE and NPRO power (second attachment).  Start date of the trends is 11-5-2016, roughly 3.5 weeks before the start of O2.

It is clear when we started adjusting the HPO diode box operating currents on 4-18-2017; previous to that date we were adjusting the currents on an as-needed basis.  The large jump in H1:PSL-OSC_DB1_PWR near the end of the trend is when we swapped that diode box for a spare in early June.  I was also going to include a trend of the HPO DB operating currents, but a read-back issue with DB3 makes this useless; the power supply reports an operating current to the PSL Beckhoff of 100 A, not the 52.3 A displayed on the front of the power supply (a power supply swap should fix this issue, planning for this as well after O2).  In light of that I will make a plot similar to Matt's here and post it as a comment.

On the 2nd attachment, it is clear the drop in the FE power coincides with the drop in the NPRO power.  This is an issue because we are currently unable to increase the FE power by adjusting the FE DB operating currents or temperatures; we suspect this is due to the low NPRO power.  It should be noted that the calibration of H1:PSL-PWR_NPRO_OUTPUT is not correct; the NPRO output power was measured back in May to be 1.36 W.  We will correct this when we swap our aging NPRO for a spare at the end of O2.