Fluent Bit

Fluent Bit is a fast and lightweight log processor and forwarder that enables collecting, parsing, filtering, and routing logs and metrics from various sources to multiple destinations. Originally developed by Eduardo Silva and now maintained by Fluent (formerly Treasure Data) as a CNCF Graduated project, Fluent Bit has over 6,500 stars as of 2026 and is designed as the high-performance, low-resource companion to Fluentd. While Fluentd is written in Ruby and targets general-purpose log aggregation with ~40MB memory usage, Fluent Bit is written in C and targets edge and constrained environments with as little as 450KB of memory, making it ideal for containerized deployments, IoT devices, and embedded systems. Key features include: 80+ input plugins (collecting from tail files, systemd journal, Docker logs, Kubernetes node and pod metrics, MQTT, statsd, TCP/UDP, exec, Windows Event Log, PostgreSQL, MySQL query results, Prometheus remote write, OpenTelemetry, and many more), 100+ output plugins (forwarding to Elasticsearch, OpenSearch, Kafka, AWS S3/Kinesis/CloudWatch, Google Cloud Logging, Azure Log Analytics, Splunk, Datadog, New Relic, Loki, InfluxDB, PostgreSQL, HTTP endpoints, and file), filter plugins (modifying records with grep, parser, modify, Lua scripting, Kubernetes metadata enrichment, and AWS metadata), parser support for common formats (JSON, regex, logfmt, LSTV, and custom patterns), buffering with in-memory and filesystem backends, retry logic with exponential backoff, TLS/SSL encryption for secure data transmission, multi-architecture support (x86, ARM, ARM64), and seamless integration with Kubernetes as a DaemonSet.