What happened:

An accidental button push took H1 out of Observe at 19:03:46UTC, and H1 was returned to Observe at 19:25:31UTC after the PSL diode chiller was filled.

Timeline Details:

• The filter, ETMY_L3_DRIVEALIGN_L2L FM3, was briefly turned off and back on, and the glitch was visible in H1 DARM.
• The filter was in a non-standard configuration approximately between 1145364642 and 1135364645 GPS time (times taken from dataview plot).
• During that time I was talking to Peter King - he called to let me know that the PSL Diode Chilller was in alarm, low water level.
• Trend of 7 day history of this alarm is attached, and shows that although the chiller was filled on the 24th, it quickly went back into alarm for low water level.
• History is that when filled, the amount added has been 125ml.
• I added 250ml today.
• This alarm should be watched closely in the next 24 hours to insure it doesn't go back into low water level alarm.
• I finished filling the chiller by 19:24:00UTC.
• I returned to the Control Room and returned H1 to Observe at 19:25:13UTC.

Given these activites, after ETMY L2L filter back was restored and resulting glitch in DARM passed, the IFO was not disturbed again until I filled the chiller between 19:19:00UTC and 19:24:00UTC.

Timeline Summary:

• 19:03:46 H1 out of Observe
• approx. 19:03:50, ETMY filter bank was back in standard config.
• H1 undisturbed for approx. 15 minutes
• 19:19:31UTC I take a "before" snapshot of the chiller warning
• I fill the chiller
• 19:23:52UTC I take the "after" snapshot of the chilller and then exit the chiller room
• 19:25:13UTC H1 is returned to Observe

IFO Guardian Log Details:

2015-12-27T08:01:00.52831 IFO [ALL_NODES_OK.enter]
2015-12-28T19:03:46.23209 IFO [ALL_NODES_OK.run] USERMSG: node NOTIFICATION: DIAG_SDF
2015-12-28T19:03:46.35568 IFO state returned jump target: DROP_OK
2015-12-28T19:03:46.35588 IFO [ALL_NODES_OK.exit]
2015-12-28T19:03:46.41705 IFO JUMP: ALL_NODES_OK->DROP_OK
2015-12-28T19:03:46.41724 IFO calculating path: DROP_OK->ALL_NODES_OK
2015-12-28T19:03:46.41748 IFO new target: WAITING_FOR_NODES_OK
2015-12-28T19:03:46.41790 IFO executing state: DROP_OK (5)
2015-12-28T19:03:46.41802 IFO [DROP_OK.enter]
2015-12-28T19:03:46.60496 IFO EDGE: DROP_OK->WAITING_FOR_NODES_OK
2015-12-28T19:03:46.60534 IFO calculating path: WAITING_FOR_NODES_OK->ALL_NODES_OK
2015-12-28T19:03:46.60572 IFO new target: ALL_NODES_OK
2015-12-28T19:03:46.60645 IFO executing state: WAITING_FOR_NODES_OK (20)
2015-12-28T19:03:46.60664 IFO [WAITING_FOR_NODES_OK.enter]
2015-12-28T19:03:47.92328 IFO EDGE: WAITING_FOR_NODES_OK->ALL_NODES_OK
2015-12-28T19:03:47.92369 IFO calculating path: ALL_NODES_OK->ALL_NODES_OK
2015-12-28T19:03:47.92443 IFO executing state: ALL_NODES_OK (100)
2015-12-28T19:03:47.92461 IFO [ALL_NODES_OK.enter]
 

Plots attached:

1. ETMY filter (approx. 3 seconds)
2. PSL Diode Chiller Low Alarm (30 minutes)
3. Both channels plotted together, showing the time in between where H1 is undistrubed, but not in Observe (23 minutes total, approx. 15 minutes of "unofficial undisturbed" time)