We provide only a single endpoint for probes (rest/v2/container/health/status). All readiness, liveness, and startup utilize the same endpoint. If the caches take too long to initialize (state transfer or reindexing), the server might be stopped externally as not responding.

We want to disambiguate the stages:

• Returns 200 when the cache container is fully started (ST & reindexing complete) for all caches. Available vs Degraded ignored.
• The Server process has started, and the Netty transport is up.