Last week, we had two on call events because of an increase of the temperature of the slave laser's crystals.
This must be due to the water flow which is somehow blocked into the laser head.
In order to try to clean it, this morning, we plugged a different chiller with filters and reversed the flow in the water circuit (around 8h30 UTC). We let it run for the whole maintenance and went back to the first chiller (around 11h10 UTC). 
We did not see any improvements in the temperature signals but this action was more preventive that curative. Indeed we did not have any temperature increase event since last Wenesday night ( log entry https://logbook.virgo-gw.eu/virgo/?r=48440 )

However, we observed a small effect on the transmission of the PMC and the IMC (plot 1). This is due to a decrease of the noise on the SL (plot 2) but looking at the signals of the SL, this decrease is not well understood (plot 3).