Kafka Connect Operations¶

Operational guidance for running Kafka Connect in production.

Monitoring¶

Key Metrics¶

Metric	Description	Alert
`connector-count`	Active connectors	≠ expected
`task-count`	Running tasks	≠ expected
`connector-startup-failure-total`	Startup failures	> 0
`source-record-poll-total`	Records polled	Anomaly
`sink-record-send-total`	Records sent	Anomaly
`offset-commit-failure-total`	Commit failures	> 0
`deadletterqueue-produce-total`	DLQ records	> 0

JMX MBeans¶

kafka.connect:type=connector-metrics,connector={name}
kafka.connect:type=connector-task-metrics,connector={name},task={id}
kafka.connect:type=source-task-metrics,connector={name},task={id}
kafka.connect:type=sink-task-metrics,connector={name},task={id}

Health Check¶

# Cluster info
curl http://connect:8083/

# List connectors
curl http://connect:8083/connectors

# Check connector status
curl http://connect:8083/connectors/my-connector/status

REST API Operations¶

Create Connector¶

curl -X POST http://connect:8083/connectors \
  -H "Content-Type: application/json" \
  -d @config.json

Update Configuration¶

curl -X PUT http://connect:8083/connectors/my-connector/config \
  -H "Content-Type: application/json" \
  -d @updated-config.json

Pause/Resume¶

# Pause
curl -X PUT http://connect:8083/connectors/my-connector/pause

# Resume
curl -X PUT http://connect:8083/connectors/my-connector/resume

Restart¶

# Restart connector
curl -X POST http://connect:8083/connectors/my-connector/restart

# Restart specific task
curl -X POST http://connect:8083/connectors/my-connector/tasks/0/restart

Delete¶

curl -X DELETE http://connect:8083/connectors/my-connector

Scaling¶

Add Workers¶

Start additional worker processes with same group.id:

# On new node
connect-distributed.sh config/connect-distributed.properties

Tasks automatically rebalance across workers.

Increase Tasks¶

curl -X PUT http://connect:8083/connectors/my-connector/config \
  -H "Content-Type: application/json" \
  -d '{
    "connector.class": "...",
    "tasks.max": "6"
  }'

Troubleshooting¶

Common Issues¶

Issue	Symptom	Solution
Task failed	`FAILED` status	Check logs, fix config, restart
Connector not starting	`UNASSIGNED` tasks	Check worker logs
High lag	Slow processing	Scale tasks/workers
Memory issues	OOM errors	Increase heap, reduce batch size

Check Task Status¶

curl http://connect:8083/connectors/my-connector/status | jq

{
  "name": "my-connector",
  "connector": {"state": "RUNNING", "worker_id": "worker1:8083"},
  "tasks": [
    {"id": 0, "state": "RUNNING", "worker_id": "worker1:8083"},
    {"id": 1, "state": "FAILED", "worker_id": "worker2:8083", "trace": "..."}
  ]
}

View Worker Logs¶

# Check Connect worker logs
tail -f /var/log/kafka-connect/connect.log | grep ERROR

Configuration Management¶

Validate Before Deploy¶

curl -X PUT http://connect:8083/connector-plugins/MyConnector/config/validate \
  -H "Content-Type: application/json" \
  -d @config.json

Export Configuration¶

curl http://connect:8083/connectors/my-connector/config > backup.json

Internal Topics¶

Topic	Purpose	Cleanup
`connect-offsets`	Source offsets	Compacted
`connect-configs`	Configurations	Compacted
`connect-status`	Status updates	Compacted

Verify topics exist with proper replication:

kafka-topics.sh --bootstrap-server kafka:9092 --describe \
  --topic connect-offsets

Kafka Connect - Connect overview
Error Handling - Error tolerance
Connectors - Connector guides