Blast Radius Analysis · ML Feature Serving Platform

Cold Start Latency

degradedContained

operational failure · isolated propagation

Blast Radius

0/21 nodes

Time to Detect

Visible within seconds of a new pod receiving traffic, as p99 latency spikes on that instance. In aggregate metrics (cluster-level p99), detectable during rolling deployments when cold pod fraction is significant.

Preventive Mitigations

Confidence

Strong

Failure Cascade

(no topology nodes mapped for this failure mode)

This failure mode is referenced in the scenario knowledge but has no connected topology nodes. Impact scope is scenario-level, all components should be considered potentially affected.

Severity at this step: degraded

Detection Signals

Latency SpikeLog Errors

Recovery time estimate: Self-resolving within 30–120 seconds as JIT compilation and cache population complete. For Lambda cold starts, resolved within the first invocation. With readiness probe warm-up, the issue is prevented entirely at the cost of slightly longer deployment time.

Mitigation Checklist(3 preventive, 2 reactive)

Implement synthetic warm-up in readiness probepreventsmedium

Before reporting ready, execute a synthetic sequence of requests that exercises the hot code paths: triggering JIT compilation, populating in-process caches, and establishing connection pool connections. Only then return 200 from the readiness endpoint. This delays traffic routing until the instance is actually warm.

Configure HikariCP to eagerly initialize connection poolpreventslow

Set minimumIdle = maximumPoolSize to establish all connections on startup rather than lazily. connectionInitSql can run a lightweight ping query on each new connection to verify the connection is live before adding it to the pool.

Use Kubernetes startup probe for slow-starting containerspreventslow

Configure a startup probe with a longer failureThreshold * periodSeconds budget (e.g., 120 seconds total). The startup probe gates the liveness and readiness probes from activating until the container has had enough time to initialize. Prevents the liveness probe from killing a slow- starting pod.

Enable class data sharing (CDS) for JVM servicesmedium

JVM Class Data Sharing and AppCDS pre-compile the class loading metadata into a shared archive that is memory-mapped on startup, reducing class loading time by 30–50%. Requires generating the archive as part of the container build process.

Apply Blue Green Deploymentmedium

Blue-green deployment pre-warms the new environment (connections, caches, JIT) before traffic shifts, eliminating cold-start latency that would otherwise occur during in-place deployments.

Affected Systems

Workloads

AI Embedding LookupAnalytics Heavy (OLAP)Read-Heavy API Backend

Blast radius analysis is derived from structured topology and failure mode knowledge. It models structural propagation patterns, not measured production behavior. Actual incident scope depends on runtime conditions, traffic, and recovery actions in place at the time of failure.