AxonOps Kafka Schema Registry Dashboard Metrics Mapping

Overview

The Schema Registry Dashboard provides monitoring of your Kafka Schema Registry service, including schema counts and types, request performance, and underlying Kafka store operations. This dashboard is available when the AxonOps agent is configured with Schema Registry metrics scraping enabled.

For setup instructions, see Schema Registry (AxonOps).

Metrics Mapping

Dashboard Metric	Description	Attributes
Schema Counts
kafka_schema_registry_registered_count_registered_count	Total number of registered schemas	dc, rack, hostname
kafka_schema_registry_deleted_count_deleted_count	Number of deleted schemas	dc, rack, hostname
kafka_schema_registry_master_slave_role_master_slave_role	Node role (1 = primary, 0 = replica)	dc, rack, hostname, host_id
Schema Types
kafka_schema_registry_avro_schemas_created_avro_schemas	Count of Avro format schemas	dc, rack, hostname
kafka_schema_registry_json_schemas_created_json_schemas	Count of JSON format schemas	dc, rack, hostname
kafka_schema_registry_protobuf_schemas_created_protobuf_schemas	Count of Protobuf format schemas	dc, rack, hostname
Request Metrics
kafka_schema_registry_jersey_metrics_request_rate	Requests per second	dc, rack, hostname
kafka_schema_registry_jersey_metrics_request_latency_95	95th percentile request latency (ms)	dc, rack, hostname
kafka_schema_registry_jersey_metrics_request_error_count	Error responses by HTTP status code	dc, rack, hostname, http_status_code
Kafka Store Operations
kafka_producer_producer_metrics_connection_count	Active producer connections to Kafka	dc, rack, hostname
kafka_consumer_consumer_metrics_connection_count	Active consumer connections to Kafka	dc, rack, hostname
kafka_producer_producer_metrics_flush_time_ns	Time to flush data to Kafka store (ns)	dc, rack, hostname, client_id
kafka_producer_producer_metrics_failed_authentication_rate	Rate of failed authentication attempts	dc, rack, hostname
kafka_consumer_consumer_fetch_manager_metrics_records_consumed_rate	Rate of records consumed from Kafka	dc, rack, hostname
kafka_consumer_consumer_fetch_manager_metrics_fetch_latency_avg	Average consumer fetch latency (ms)	dc, rack, hostname
kafka_consumer_consumer_fetch_manager_metrics_records_lag	Consumer lag in records	dc, rack, hostname

Query Examples

Schema Counts

// Total registered schemas
kafka_schema_registry_registered_count_registered_count{dc=~'$dc',rack=~'$rack',hostname=~'$hostname'}

// Deleted schemas
kafka_schema_registry_deleted_count_deleted_count{dc=~'$dc',rack=~'$rack',hostname=~'$hostname'}

// Primary vs replica role
kafka_schema_registry_master_slave_role_master_slave_role{dc=~'$dc',rack=~'$rack',hostname=~'$hostname'}

Schema Types

// Avro schemas
kafka_schema_registry_avro_schemas_created_avro_schemas{dc=~'$dc',rack=~'$rack',hostname=~'$hostname'}

// JSON schemas
kafka_schema_registry_json_schemas_created_json_schemas{dc=~'$dc',rack=~'$rack',hostname=~'$hostname'}

// Protobuf schemas
kafka_schema_registry_protobuf_schemas_created_protobuf_schemas{dc=~'$dc',rack=~'$rack',hostname=~'$hostname'}

Request Performance

// Request rate
kafka_schema_registry_jersey_metrics_request_rate{dc=~'$dc',rack=~'$rack',hostname=~'$hostname'}

// Request latency (p95)
kafka_schema_registry_jersey_metrics_request_latency_95{dc=~'$dc',rack=~'$rack',hostname=~'$hostname'}

// Error counts by HTTP status code
kafka_schema_registry_jersey_metrics_request_error_count{dc=~'$dc',rack=~'$rack',hostname=~'$hostname'}

Kafka Store Operations

// Total connections (producers + consumers)
sum(kafka_producer_producer_metrics_connection_count{dc=~'$dc',rack=~'$rack',hostname=~'$hostname'})
sum(kafka_consumer_consumer_metrics_connection_count{dc=~'$dc',rack=~'$rack',hostname=~'$hostname'})

// Store flush time
kafka_producer_producer_metrics_flush_time_ns{dc=~'$dc',rack=~'$rack',hostname=~'$hostname'}

// Failed authentication rate
kafka_producer_producer_metrics_failed_authentication_rate{dc=~'$dc',rack=~'$rack',hostname=~'$hostname'}

// Consumer records consumed rate
sum(kafka_consumer_consumer_fetch_manager_metrics_records_consumed_rate{dc=~'$dc',rack=~'$rack',hostname=~'$hostname'})

// Consumer fetch latency
avg(kafka_consumer_consumer_fetch_manager_metrics_fetch_latency_avg{dc=~'$dc',rack=~'$rack',hostname=~'$hostname'})

// Consumer lag
sum(kafka_consumer_consumer_fetch_manager_metrics_records_lag{dc=~'$dc',rack=~'$rack',hostname=~'$hostname'})

Panel Organization

Schema Counts

• Total registered schemas
• Deleted schemas count
• Master vs Slave role

Schema Types

• Avro schemas
• JSON schemas
• Protobuf schemas

Requests

• Request latency (p95)
• Request rate
• Error codes (by HTTP status)

Kafka Store Operations

• Failed authentication rate
• Store flush time (ns)
• Connection count (producers and consumers)
• Consumer fetch latency (avg)
• Consumer records consumed rate
• Consumer lag

Filters

• dc: Filter by datacenter

• rack: Filter by rack

• hostname: Filter by specific Schema Registry node

Best Practices

Schema Growth Monitoring

• Track registered schema count over time for capacity planning
• A rising deleted count may indicate schema churn or cleanup activity
• Monitor schema type distribution to understand format adoption

Request Performance

• p95 latency should remain low (typically under 50ms)
• Rising error counts may indicate compatibility violations or client issues
• Monitor request rate to understand API usage patterns

Kafka Store Health

• Consumer lag should remain near zero for a healthy registry
• High flush times may indicate Kafka broker issues
• Failed authentication rate should be zero under normal operation
• Connection count changes may indicate registry restarts or scaling events

Troubleshooting

• No metrics appearing: Verify SCHEMA_REGISTRY_METRICS_URL is configured and reachable
• High consumer lag: Check Kafka broker health and Schema Registry logs
• Rising error counts: Review Schema Registry logs for compatibility or serialization errors