Erik, Camilla
This was already installed on ETMY and the code described by Huy-Tuong Cao in 72229. Today Erik installed conda on h1hwsex and we got the new code running. I had to take a new ETMX reference while the IFO was hot (old py2 pickle files have a different format/encoding to py3). So next Tuesday I should re-take this reference with a cold IFO. This update can be done to the ITMs tomorrow.
To install conda or add it to the path, Erik ran command: '/ligo/cds/lho/h1/anaconda/anaconda2/bin/conda init bash'
Then after closing/reopening a terminal I can pull the hws-server/-/tree/fix/python3 code:
Stop and restart the code after running 'conda activate hws' to use the correct python paths.
Had a few errors that needed to be fixed before the code ran successfully:
- allow_pickle=True, encoding='latin1' to the numpy.load line 165 in run_HWS.py
- fix indentation errors (tabs to spaces) and print statements in /home/controls/HWS_code_temporary_home.bkp/common/hws_code/utilities/stream_HWS_intensity.py lines 106,287,315,321, /home/controls/hws/HWS/HS_Centroids.py line 273 and /home/controls/hws/HWS/HS_Image.py line 585
- Took new references as couldn't unpickle the old py2 references, I'm sure they could have been a work around if we wanted.
New code now running on all optics. Took new references on ITMs and ETMX.
New ETMX reference taken at 16:21UTC with ALS shuttered, after IFO had been down for 90 minutes. Installed at ITMX, new reference taken at 16:18UTC. Installed at ITMY, new reference taken at 16:15UTC. realized where I previously edited code in /home/controls/hws/HWS/ I should have just pulled the new fix/python3 HWS code I went back did this for ITMs and ETMX.
ITMX SLED power has quickly decayed, see attached, so I adjusted the frame rate form 5Hz to 1Hz, instructions in TCS wiki page. This SLED was recently replaced 71476 and was a 2021 SLED so not particularly old. LLO has seen similar issues 66713 but over hours rather than weeks. We need to deicide what to do about this, we could try to touch the trimpoint or contact the company. ITMY SLED is fine.
There was two separate errors on ITMY that stopped the code in the first few minutes, a "segmentation fault" and a "fatal IO error 25". We should watch that this code continues the run without issues.
ITMY code keeps stopping and has been stopped for last few hours. I cannot access the computer, maybe it has crashed? There is an orange light on h1hwsmsr.
ITMX spherical power is very noisy. This appears to be because the new ITMX reference has a lot of dead pixels, I've noted them but haven't yet added them to the dead pixels files as cannot edit the read-only file- TCS wiki link.
ITMY- TJ and I restarted h1hwsmrs1 in the MSR and I restarted the code. We were getting a regular "out of bounds" error, attached, but the data now seems to be running fine. The fix/python3 code didn't have all the master commits in it so when we restarted mrs1 some of the channels were not restarted, found by Dave in 73013. I've updated the fix/python3 code and we should pull this commit and kill/restart the ioc and hws itmy code tomorrow.
ITMX - the bad pixels were added and the data is now mcuh cleaner, updated insruciton in the TCS wiki.
Pulled new python3 code with all channels to both ITM and ETMX computers (already on ETMY). Killed and restart softIoc on h1hwsmsr1 for ITMY following instructions in 65966. The 73013 channels are now running again.
Took new references on all optics at 20:20UTC after the IFO had been unlocked and CO2 lasers off for 5 hours. RH settings nominal IX 0.4W, IY 0.0W, EX 1.0W, EY 1.0W.