According to the excitation flag shown in Fig.1, a script seemed to work fine if an injection around 15:00 JST on 11/23 Sat. (t=15h in Fig.1). is a human activity not due to the script.
We made lots of injections for a debug of the script during 18:00-21:00 JST on 11/22 Fri. (t=[-6h, -3h]). And the, lots of injection with IFO lock were done in during 0:00-5:00 JST on 11/23 Sat. (t=[0h, 5h]). This behavior follows our expectation. After then, injection wasn't done except around 15:00 JST on 11/23 Sat. (t=15h). If it's due to this script, The fact that there was no injection during IFO lock around t=6h, 10h and 12h is strange. So I believe that the injection on Saturday evening should be a human activity though I cannot find a related klog, and the script properly suspended injections around 5:00 JST on 11/23 Sat. And also, there was no injection around 7:00 JST on 11/24 Sun. (t=1d,7h). I guessed that injections around 7:40 JST on 11/24 Sun. were done after stopping the script though I cannot find a related klog again.
Anyway, Because what mentioned above includes some imagines, we need to gather information about more detailed timeline of weekend activities in order to confirm that it's not really a bug of script or not.