ASG update command fails if ASG was set to 0,0,0
We just had a case where a pipeline for a client was run where one of the asg's involved (the inactive one) was set to scaling policy of 0,0,0. When the asg update command is run, it scales down to 0.0.0 first, then back to whatever it was before (https://gitlab.olindata.com/olindata/akinaka/blob/master/akinaka/update/asg/update_asg.py#L101). After that it waits for instances to come back up. However, if the old setting was for some reason 0,0,0 it will wait indefinitely.