To the best of my knowledge, if for example the scheduled event returns a "null pointer", the scheduled event will not run again untill the next reboot of the application. You could create a submicroflow in the scheduled event flow to catch the error with error handling, not only will you catch the error, the scheduled event will also continue to run.
As indicated by Roy and Ronald, subsequent events may not fire, also forother scheduled events. Over time, your app might not process events at all anymore. It seems the thread gets lots. Only a restart fixes this.
So, always, use custom error handling on your scheduled event main microflow, log the error, with the tickbox to log the stack trace.
Where necessary, add logic to report issues in your runs to admins, on dashboard, by email/SMS/???