Goal: The health actuators of our microservices include ping tests to external services. Naturally, these ping requests sometimes time out or fail for other reasons, which leads to a DOWN response by the /health actuator. We would like SBA Server to ignore 1-2 of those DOWN requests and keep the status for the service in UP, until let’s say 3 or more DOWN responses were received.

I studied all the available configuration options for SBA Server but to be honest, I am confused how they all play together, and how I can achieve this goal. I naively thought the default-retries setting would accomplish this, but it didn’t. I assume (hope) the desired behavior can be achieved by properly concerting multiple of those values, as it doesn’t seem there is a single one that would achieve this goal.

For reference, these are the configuration options that I think may apply here:

• spring.boot.admin.monitor.status-interval: Time interval to check the status of instances.
• spring.boot.admin.monitor.status-lifetime: Lifetime of status. The status won’t be updated as long the last status isn’t expired.
• spring.boot.admin.monitor.default-timeout: Default timeout when making requests.
• spring.boot.admin.monitor.default-retries: Default number of retries for failed requests.

Any thoughts?