Key features
- 500+ plugins — input, filter, output plugins for every APAC log source and destination in the ecosystem
- Kubernetes DaemonSet — node-level APAC container log collection with Kubernetes metadata enrichment
- JSON-native model — unified structured log events for APAC log querying and routing
- Output buffering — disk/memory buffering with retry for reliable APAC log delivery during outages
- Fluent Bit integration — lightweight C-native APAC edge collector feeding Fluentd aggregator tier
- Multi-destination routing — fan-out APAC logs to Elasticsearch, S3, BigQuery simultaneously
- CNCF graduation — production-ready for APAC enterprise log pipelines with Ruby plugin ecosystem
Best for
- APAC platform engineering teams building centralised Kubernetes log aggregation pipelines that need to route logs to multiple APAC destinations (Elasticsearch for search, S3 for archive, Slack for alerts) with a single log collection DaemonSet
- Engineering organisations migrating APAC log infrastructure from legacy centralized syslog to cloud-native log routing, needing Fluentd's extensive plugin ecosystem to support both legacy on-premise log sources and Kubernetes container logs
- APAC regulated-industry platform teams that need structured JSON log events for compliance audit logging, where Fluentd's parser plugins normalise heterogeneous APAC application log formats to a queryable structured schema
- Platform teams building APAC two-tier log pipelines with Fluent Bit (lightweight DaemonSet collector) forwarding to Fluentd aggregators (heavy parsing, enrichment, multi-destination routing) for large APAC Kubernetes clusters with memory-constrained nodes
Limitations to know
- ! Ruby runtime overhead — Fluentd is written in Ruby and has higher per-pod memory footprint (50-100MB) than Fluent Bit (5-15MB); APAC large-scale DaemonSet deployments on thousands of nodes should evaluate Fluent Bit as the primary collector with Fluentd as the aggregation tier
- ! Plugin quality variance — Fluentd's 500+ plugin ecosystem includes community-maintained plugins with varying maintenance quality; APAC teams should audit plugin release frequency and issue tracker activity for specific output plugins before production adoption
- ! Configuration complexity at scale — Fluentd's Ruby DSL configuration (fluent.conf) becomes complex in large APAC deployments with many sources, filters, and outputs; APAC teams should use configuration management (Helm charts, Kustomize) and test configuration changes in staging before production rollout
- ! No built-in log storage — Fluentd routes logs but does not store them; APAC platform teams must provision and maintain an appropriate log storage backend (Elasticsearch, Loki, S3) separately, adding infrastructure operational scope beyond Fluentd itself
About Fluentd
Fluentd is a CNCF-graduated open-source unified logging layer that provides APAC platform engineering teams with a flexible, plugin-based log collection, transformation, and routing pipeline — collecting logs from Kubernetes pod stdout/stderr (via DaemonSet), application log files, system journal, and other APAC infrastructure sources, processing them through filter and parser plugins, and routing them to any supported output destination (Elasticsearch/OpenSearch, Amazon S3, Google BigQuery, Azure Monitor, Splunk, Datadog, Kafka, and 500+ additional destinations via the plugin ecosystem).
Fluentd's unified JSON log model — where all logs entering Fluentd's pipeline are normalised to a structured JSON event record regardless of the original log format (Apache access logs, Nginx logs, application-specific log formats, syslog) using Fluentd's parser plugins (multiline, grok, regex, JSON) — enables APAC platform engineering teams to query, filter, and route structured log data across APAC infrastructure without maintaining separate log parsing logic for each application log format, enabling Elasticsearch log queries using structured JSON fields rather than full-text regex patterns.
Fluentd's DaemonSet log collection model — where Fluentd runs as a Kubernetes DaemonSet with one pod per APAC cluster node, mounting the node's `/var/log/containers/` directory and tailing Kubernetes container log files using the `in_tail` plugin with Kubernetes metadata enrichment (pod name, namespace, labels, deployment) via the `filter_kubernetes_metadata` plugin — enables APAC platform teams to collect all Kubernetes pod logs centrally without modifying application containers or requiring application-level log shipping libraries.
Fluentd's output buffering and reliability — where Fluentd buffers log events to disk or memory before forwarding to APAC log destinations, with configurable retry logic (exponential backoff, maximum retry count, retry timeout) and a persistent buffer that survives Fluentd pod restarts — enables APAC platform engineering teams to build reliable log pipelines that do not lose APAC log events during downstream destination outages, log destination maintenance windows, or APAC network connectivity interruptions between the cluster and the log destination.
Fluent Bit's relationship to Fluentd — where Fluent Bit is a lightweight log collector written in C that is 20-40x more memory-efficient than Fluentd, typically deployed as the APAC edge DaemonSet collector with Fluentd as the APAC central aggregation tier receiving Fluent Bit output — enables APAC platform teams to build a two-tier log pipeline where memory-constrained APAC Kubernetes nodes run Fluent Bit and a smaller number of Fluentd aggregator pods handle parsing, enrichment, and routing to APAC log destinations.
Beyond this tool
Where this category meets practice depth.
A tool only matters in context. Browse the service pillars that operationalise it, the industries where it ships, and the Asian markets where AIMenta runs adoption programs.
Other service pillars
By industry