Regarding detecting stalled slapos node instance in the future.
Also, Romain suggested that "slapos node instance" command would kill the previous run if it has been running for more than 24h.
No. I suggested that slapos node instance reports a problem to the master (ie, to create a ticket) so that an admin could check.
This is just a first step towards implementing such a feature. The big question is how to report a stall to the master; ideas and suggestions are welcome.