The HealthCheck service is a mechanism provided by CycleCloud to terminate any instances in an undesired state. The most common of these is when a node fails to converge after repeated attempts. Unfortunately, this can also make debugging cluster-init and custom Chef scripts difficult, since it can terminate the instance while you're still trying to troubleshoot the problem.


To disable HealthCheck on a node, run the jetpack keepalive <timeframe> command. On Linux instances, you can specify a timeframe in hours or forever. On Windows instances, only the forever argument is supported.