At the moment if the connection with the headstage (?) is lost the bonsai workflow crashes.
This is a critical failure mode because if nobody notices this hours of recordings can be lost plus the commutator would not spin, leading to tangling. This happened twice in the last 48h. It is important to notice that after the workflow restarts, the recording is stable for several hours, so it is really a point event where something strange happens rather than the overall connection becoming loose.
We can improve the physical connection between the probe and the headstage and the headstage and the neuropixels cable, however this will never fully eliminate the problem.
Would there be a way to solve this in software to automatically restart the workflow if this happen (plus keeping the alert so user is aware)?
At the moment if the connection with the headstage (?) is lost the bonsai workflow crashes. This is a critical failure mode because if nobody notices this hours of recordings can be lost plus the commutator would not spin, leading to tangling. This happened twice in the last 48h. It is important to notice that after the workflow restarts, the recording is stable for several hours, so it is really a point event where something strange happens rather than the overall connection becoming loose.
We can improve the physical connection between the probe and the headstage and the headstage and the neuropixels cable, however this will never fully eliminate the problem.
Would there be a way to solve this in software to automatically restart the workflow if this happen (plus keeping the alert so user is aware)?