Is your feature request related to a problem? Please describe.
In the stoppable check API, if an instance is not alive, Helix still attempts to run the custom instance and partition checks. However, because the instance’s health endpoint is unavailable, these checks always fail. This creates a maintenance issue, as a successful stoppable check is required before we can safely restart the host.
Describe the solution you'd like
Ideally, if the user chooses to skip the Helix-level “INSTANCE_NOT_ALIVE” check, the instance’s aliveness becomes irrelevant, and the custom checks do not need to be performed. This would allow the stoppable check to pass, enabling the host to be restarted without unnecessary failures.
Additional context
N/A
Is your feature request related to a problem? Please describe.
In the stoppable check API, if an instance is not alive, Helix still attempts to run the custom instance and partition checks. However, because the instance’s health endpoint is unavailable, these checks always fail. This creates a maintenance issue, as a successful stoppable check is required before we can safely restart the host.
Describe the solution you'd like
Ideally, if the user chooses to skip the Helix-level “INSTANCE_NOT_ALIVE” check, the instance’s aliveness becomes irrelevant, and the custom checks do not need to be performed. This would allow the stoppable check to pass, enabling the host to be restarted without unnecessary failures.
Additional context
N/A