If your Kafka cluster enforce ACLs see ACLs.md.
-
You need a compacted topic (configured via
RIPT_STATE_TOPIC) for RIPT's state storage, RIPT can create this topic if allowed, otherwise provision it with a lower segment.ms, 86400000 (1 day) is recommended -
Decide on a good partition count for RIPT's state topic which will be used as sharding for large clusters
-
Choose an interval for tracking the offsets - A daily scan should be okay for most of the cases, for more dynamic environments choose a lower interval
Environment variables (see .env.example) should be self explanatory
Create/edit .env:
RIPT_KAFKA_BROKERS=localhost:9092 # Your Kafka brokers (comma-separated)
RIPT_SCAN_INTERVAL_MINUTES=5 # How often to scan
RIPT_STATE_TOPIC=ript-state # Internal state topic (auto-created if permitted)
RIPT_HTTP_PORT=8080 # Dashboard port
RIPT_LOG_LEVEL=info # debug|info|warn|error