Took a quick look at the Diode chiller.  At approximately 16:22 UTC (8:22 PST) last Friday, November 22nd, the flow reported by the Diode Chiller dropped slightly (~19.7 lpm down to ~19.2 lpm) and then begain behving somewhat erratically (dropped as low as ~17.2 lpm, reported flow stayed between 17.2 lpm and 18.4 lpm); this behavior continued over the weekend and into this morning, with the reported flow returning to ~19.1 lpm at ~17:00 UTC (9:00 PST).  The other diode chiller signals included in Ed's plot show no other reason to think something is going wrong with the chiller (conductivity behaving as normal, water temp is steady).  I checked on the chiller itself and it was operating normally, no odd or out-of-place noises or vibrations.  This is still plenty of flow to cool the diode boxes, so laser performance was not affected (as can be seen in the attachment, diode box heat sink temp (what the diode chiller actually cools, the pump diodes are cooled by TECs attached to this heat sink) doesn't change in response to this reported flow change).

Therefore, at this point I have no reason to suspect a problem with the Diode Chiller.  This could be a one-off event, or an early sign of flow meter failure (each chiller has an internal flow meter for monitoring flow).  Recall that this chiller was borrowed from 3IFO (since replaced with a repaired and working unit to prevent interruption of PSL operation during pre-O3 commissioning) during an emergency in February 2019 where both our in-service and backup chillers experienced failures (many alogs, but the most relelvent is here).  As noted in the linked alog, this chiller is still running the old style impeller-based internal flow meters, which have a shorter lifespan than the newer vortex flow meters we installed in the PSL chillers (sans 3IFO) in 2015.  Therefore I'll keep an eye on this for signs of a failing flow meter.