Danny, Sheila
Concerning the A2L script, I could not reproduce the problem. My best guess is: sometimes the online NDS connection returns bad data for the first second. So I modified the script so that it discards the first chunk of data it receives.
I also added a check on the arm transmitted power: if LSC-TR_X_NORM_IN is lower than 1000, then the script stops the measurement, since the IFO is likely unlocked.
If the script fails again returning NaNs, you can simply try to run it again.