|
| 1 | +# Kafka-Elasticsearch Indexer Configuration File |
| 2 | +# |
| 3 | +# This file configures the kafka-es-indexer service that consumes OpenCue |
| 4 | +# monitoring events from Kafka and indexes them into Elasticsearch for |
| 5 | +# historical analysis and querying. |
| 6 | +# |
| 7 | +# Data Flow: Cuebot (Producer) -> Kafka -> kafka-es-indexer (Consumer) -> Elasticsearch |
| 8 | + |
| 9 | +# ============================================================================= |
| 10 | +# KAFKA CONFIGURATION |
| 11 | +# ============================================================================= |
| 12 | +kafka: |
| 13 | + # Kafka bootstrap servers (comma-separated list) |
| 14 | + # Multiple brokers can be specified for high availability |
| 15 | + # Default: localhost:9092 |
| 16 | + bootstrap_servers: "localhost:9092" |
| 17 | + |
| 18 | + # Consumer group ID |
| 19 | + # All indexer instances with the same group_id will share partition |
| 20 | + # assignments and coordinate offset commits. Use a unique ID per cluster. |
| 21 | + # Default: opencue-elasticsearch-indexer |
| 22 | + group_id: "opencue-elasticsearch-indexer" |
| 23 | + |
| 24 | + # What to do when there is no initial offset in Kafka |
| 25 | + # Options: |
| 26 | + # earliest - Start from the oldest available message |
| 27 | + # latest - Start from the newest message (skip historical) |
| 28 | + # Default: earliest |
| 29 | + auto_offset_reset: "earliest" |
| 30 | + |
| 31 | + # Enable automatic offset commits |
| 32 | + # When true, offsets are committed periodically based on auto_commit_interval_ms |
| 33 | + # When false, offsets are committed manually after each message is processed |
| 34 | + # Default: true |
| 35 | + enable_auto_commit: true |
| 36 | + |
| 37 | + # Interval between automatic offset commits (in milliseconds) |
| 38 | + # Only used when enable_auto_commit is true |
| 39 | + # Lower values reduce duplicate processing on restart but increase overhead |
| 40 | + # Default: 5000 (5 seconds) |
| 41 | + auto_commit_interval_ms: 5000 |
| 42 | + |
| 43 | + # Maximum number of records to fetch per poll |
| 44 | + # Higher values improve throughput but increase memory usage |
| 45 | + # Default: 500 |
| 46 | + max_poll_records: 500 |
| 47 | + |
| 48 | + # Kafka session timeout (in milliseconds) |
| 49 | + # If the consumer doesn't send heartbeats within this interval, |
| 50 | + # it will be removed from the consumer group and partitions will be rebalanced |
| 51 | + # Default: 30000 (30 seconds) |
| 52 | + session_timeout_ms: 30000 |
| 53 | + |
| 54 | + # Kafka topics to subscribe to |
| 55 | + # These are the event topics published by Cuebot |
| 56 | + # Default: all OpenCue event topics |
| 57 | + topics: |
| 58 | + - "opencue.job.events" |
| 59 | + - "opencue.layer.events" |
| 60 | + - "opencue.frame.events" |
| 61 | + - "opencue.host.events" |
| 62 | + - "opencue.proc.events" |
| 63 | + |
| 64 | +# ============================================================================= |
| 65 | +# ELASTICSEARCH CONFIGURATION |
| 66 | +# ============================================================================= |
| 67 | +elasticsearch: |
| 68 | + # Elasticsearch URL |
| 69 | + # Can be a single node or a load balancer in front of a cluster |
| 70 | + # Default: http://localhost:9200 |
| 71 | + url: "http://localhost:9200" |
| 72 | + |
| 73 | + # Username for Elasticsearch authentication (optional) |
| 74 | + # Required when Elasticsearch has security features enabled |
| 75 | + # Can also be set via ELASTICSEARCH_USERNAME environment variable |
| 76 | + # username: "elastic" |
| 77 | + |
| 78 | + # Password for Elasticsearch authentication (optional) |
| 79 | + # Required when Elasticsearch has security features enabled |
| 80 | + # Can also be set via ELASTICSEARCH_PASSWORD environment variable |
| 81 | + # password: "changeme" |
| 82 | + |
| 83 | + # Index name prefix for all OpenCue event indices |
| 84 | + # Indices are created with pattern: {prefix}-{event-type}-{date} |
| 85 | + # Example: opencue-frame-events-2024.11.29 |
| 86 | + # Default: opencue |
| 87 | + index_prefix: "opencue" |
| 88 | + |
| 89 | + # Number of primary shards for event indices |
| 90 | + # More shards allow parallel indexing and searching |
| 91 | + # For small deployments, 1 shard is sufficient |
| 92 | + # For large deployments with many events, consider 3-5 shards |
| 93 | + # Default: 1 |
| 94 | + num_shards: 1 |
| 95 | + |
| 96 | + # Number of replica shards for event indices |
| 97 | + # Replicas provide redundancy and improve read throughput |
| 98 | + # Set to 0 for development/testing, 1+ for production |
| 99 | + # Default: 0 |
| 100 | + num_replicas: 0 |
| 101 | + |
| 102 | + # Maximum number of events to batch before sending to Elasticsearch |
| 103 | + # Higher values improve throughput but increase latency and memory usage |
| 104 | + # Events are also flushed based on flush_interval_ms |
| 105 | + # Default: 100 |
| 106 | + bulk_size: 100 |
| 107 | + |
| 108 | + # Maximum time to wait before flushing events to Elasticsearch (in milliseconds) |
| 109 | + # Events are flushed when either bulk_size is reached or this interval elapses |
| 110 | + # Lower values reduce latency but increase indexing overhead |
| 111 | + # Default: 5000 (5 seconds) |
| 112 | + flush_interval_ms: 5000 |
| 113 | + |
| 114 | +# ============================================================================= |
| 115 | +# LOGGING CONFIGURATION |
| 116 | +# ============================================================================= |
| 117 | +# |
| 118 | +# Logging is configured via the LOG_LEVEL environment variable or --log-level |
| 119 | +# CLI argument. This section documents the available options. |
| 120 | +# |
| 121 | +# Options: trace, debug, info, warn, error |
| 122 | +# Default: info |
| 123 | +# |
| 124 | +# Examples: |
| 125 | +# - trace: Very verbose, includes all debug info (development only) |
| 126 | +# - debug: Detailed info including each received event |
| 127 | +# - info: Standard operation logs (recommended for production) |
| 128 | +# - warn: Only warnings and errors |
| 129 | +# - error: Only errors |
| 130 | +# |
| 131 | +# Can also use RUST_LOG env var for fine-grained control: |
| 132 | +# RUST_LOG=kafka_es_indexer=debug,rdkafka=info |
0 commit comments