WP11293 Increase rate of LSC and ASCIMC channels
Daniel, Dave:
Daniel created new h1lsc and h1ascimc models which increased the DAQ rate of 3 DQ channels from 2kHz to 16kHz.
h1lsc:
H1:LSC-REFL_A_RF9_I_ERR_DQ datarate change from 2048 to 16384
h1ascimc:
H1:IMC-WFS_A_DC_PIT_OUT_DQ datarate change from 2048 to 16384
H1:IMC-WFS_A_DC_YAW_OUT_DQ datarate change from 2048 to 16384
I did not restart these models to load the new DAQ configuration, instead I pressed the "DAQ LOAD" button on the models' GDS_TP screen. A DAQ restart was required.
WP11299 Restart h1isietmx to test new SDF safe.snap
Jim, TJ, EJ, Dave:
The h1isietmx model was restarted using Jim's new safe.snap file. The original safe.snap was missing some channels, which was also the case for h1isiitmx and was discovered last Sunday's model restart.
The new safe.snap was found to be good.
One point of interest is that the SDF MEDM reports 3047 set point channels, and the safe.snap file has 3506 channels in it. EJ explained that for each filtermodule, the SDF combines the two switch channels (SW1S, SW2S) into one (SWSTAT). There are 459 filter modules in this model, and the maths works out 3047+459=3506.
WP11290 h1hwinj1 revert LALAPPS back to original version
Erik, Jonathan, Dave:
This turned out to be much more involved than originally thought due to dependencies. We were not able to perform the reversion during maintenance.
I did restart psinject on h1hwinj1 during maintenance to reset the memory usage to 7%.
WP11304 Temporarily remove INJ_CW GAIN/TRAMP channels from h1calinj SDF
Dave:
As a temporary measure while we are running the memory-leaking LALPULSAR, I stopped the monitoring of the INJ_CAL_GAIN and INJ_CAL_TRAMP from h1calinj's SDF.
This will give us the ability to restart psinject and not knock H1 out of observe while doing so.
WP11291 Update tzdata on h1guardian1 to extend leapseconds officially
Jamie, Erik, Dave:
The tzdata package was updated on h1guardian1, which installed the "official" /usr/share/zoneinfo/leapseconds file and replace my hand-edited one.
The only issue was that apt-get recognized there was a local edit and did not overwrite it, but neither did it report this or give us the option to overwrite it.
The solution was to restore the original file, and then "apt-get reinstall tzdata" to force the install.
Model and DAQ restarts, model reloads
Dave:
Tue11Jul2023
LOC TIME HOSTNAME MODEL/REBOOT
09:38:28 h1seiex h1isietmx
11:24:58 h1lsc DAQ LOAD
11:25:24 h1ascimc DAQ LOAD
11:27:34 h1daqdc0 [DAQ] <<< DAQ Restart for h1lsc and h1ascimc INI changes. 0-leg
11:27:45 h1daqfw0 [DAQ]
11:27:45 h1daqtw0 [DAQ]
11:27:46 h1daqnds0 [DAQ]
11:27:54 h1daqgds0 [DAQ]
11:31:49 h1daqdc1 [DAQ] <<< 1-leg
11:31:57 h1daqfw1 [DAQ]
11:31:58 h1daqnds1 [DAQ]
11:31:58 h1daqtw1 [DAQ]
11:32:06 h1daqgds1 [DAQ]
11:32:54 h1daqgds1 [DAQ] <<< GDS1 second restart
Late restart of opslogin0
Erik, Jonathan, Dave:
We did a late restart of opslogin0 at 15:02 to clear some memory and do a minor OS upgrade.
Three items visible on the CDS Overview went down during the reboot:
1. CDS-HW IOC stopped and automatically restarted
2. WAP IOC stopped and was manually restarted
3. CDS_NOMACHINE_IOC stopped and did not get automatically restarted. We had not seen this fail before, it was manually restarted today and Erik is investigating.
2+3 caused the EDC to disconnect from 20 channels.
H1 was in observe throughout.
DAQ CHANGES:
[channame, bytes-per-sample, data-rate-Hz]
was:
H1:IMC-WFS_A_DC_PIT_OUT_DQ 4 2048
H1:IMC-WFS_A_DC_YAW_OUT_DQ 4 2048
H1:LSC-REFL_A_RF9_I_ERR_DQ 4 2048
is now:
H1:IMC-WFS_A_DC_PIT_OUT_DQ 4 16384
H1:IMC-WFS_A_DC_YAW_OUT_DQ 4 16384
H1:LSC-REFL_A_RF9_I_ERR_DQ 4 16384