DEV.co
Open-Source Observability · alibaba

loongcollector

LoongCollector is a high-performance observability data collector built by Alibaba for logs, metrics, traces, and events in cloud-native environments. It claims 10x throughput advantage and 80% lower resource usage versus competitors like Fluent Bit and Vector, backed by 15+ years of Alibaba production deployment.

Source: GitHub — github.com/alibaba/loongcollector
2.2k
GitHub stars
437
Forks
C++
Primary language
Apache-2.0
License (OSI-approved)

Key facts

Objective fields from the source. Values we can't verify are shown as “Unknown” rather than guessed.

FieldValue
Repositoryalibaba/loongcollector
Owneralibaba
Primary languageC++
LicenseApache-2.0 — OSI-approved
Stars2.2k
Forks437
Open issues149
Latest releasev3.3.5 (2026-06-01)
Last updated2026-07-07
Sourcehttps://github.com/alibaba/loongcollector

What loongcollector is

C++ agent with pluggable architecture (100+ plugins, multi-language support), eBPF-based network monitoring, lock-free event pools, zero-copy serialization, and multi-tenant pipeline isolation with adaptive concurrency control. Supports remote configuration via SLS console, Kubernetes Operator, and includes SPL engine for data processing.

Quickstart

Get the loongcollector source

Clone the repository and explore it locally.

terminalbash
git clone https://github.com/alibaba/loongcollector.gitcd loongcollector# follow the project's README for install & configuration

Need it deployed, integrated, or customized instead? DEV.co ships production installs.

Best use cases

High-throughput log collection at scale

Organizations processing 10+ MB/s of observability data where CPU and memory efficiency directly impact infrastructure costs. Suitable for Kubernetes environments with hundreds of nodes or container workloads.

Unified observability stack (logs + metrics + traces + events)

Teams requiring a single agent for multi-signal collection rather than managing separate tools. Particularly valuable when integrated with Alibaba Cloud SLS or LoongSuite ecosystem.

eBPF-powered network and security monitoring

Deployments needing network flow analysis, security event collection, and kernel-level observability without application instrumentation overhead.

Implementation considerations

  • Build requires Docker and Go 1.23+; source compilation adds operational complexity vs. pre-built binaries. Docker image available but not pre-built for all platforms documented.
  • eBPF capabilities require kernel 4.4+ and elevated privileges; validate host OS and security policies before deployment at scale.
  • 149 open issues suggest active development; evaluate breaking changes and plugin stability before pinning versions in production.
  • SPL engine and 100+ plugins create a steep learning curve; operator training and plugin customization may require Alibaba documentation or community forums.
  • Remote configuration via K8s Operator, SLS console, or SDK adds operational dependency; ensure management plane availability and backup configuration strategy.

When to avoid it — and what to weigh

  • Requirement for non-Alibaba ecosystem integration — LoongCollector appears optimized for Alibaba Cloud SLS and LoongSuite agents. Integration maturity with Datadog, New Relic, Splunk, or other third-party observability platforms is not documented.
  • Limited resource environments (edge, IoT) — While lightweight, C++ agent adds complexity vs. pure Go alternatives. Deployment on severely constrained devices or heterogeneous architectures may face friction.
  • Mature, battle-tested support requirements — Project is relatively young in open-source (created Nov 2021). While Alibaba backing implies production readiness, community adoption and third-party support ecosystem remain nascent vs. Fluent Bit or Logstash.
  • Simple, minimal-config observability needs — Organizations needing a lightweight sidecar for a single log file or basic metric export may find 100+ plugins and complex pipeline architecture unnecessary overhead.

License & commercial use

Apache License 2.0 (Apache-2.0), a permissive OSI-approved open-source license allowing commercial use, modification, and redistribution with attribution and liability disclaimers.

Apache-2.0 permits commercial use without royalties or licensing fees. However, verify that your intended use (e.g., offering LoongCollector as a managed service or bundling with proprietary software) aligns with license terms. No explicit commercial support model or SLA documented in provided data; Alibaba support availability unknown.

DEV.co evaluation signals

Editorial assessment — not user reviews. Directional, with an explicit confidence level.

SignalAssessment
MaintenanceActive
DocumentationAdequate
License clarityClear
Deployment complexityModerate
DEV.co fitGood
Assessment confidenceHigh
Security considerations

eBPF-based monitoring requires kernel access and elevated privileges; audit host kernel for known vulnerabilities and restrict agent privileges to least necessary. Agent runs with access to /var/run and entire filesystem (/ mounted read-only in Docker example); validate container security context and network policies. No explicit security audit, vulnerability disclosure process, or cryptographic integrity verification mentioned in provided data. C++ implementation introduces memory safety attack surface; verify Coverity Scan coverage (badge present but details unknown) before production deployment.

Alternatives to consider

Fluent Bit

Lightweight, pure C, mature ecosystem (Kubernetes, CloudWatch, Datadog, Splunk). Slower throughput (~36 MB/s single-line vs LoongCollector's 546 MB/s) but simpler deployment and wider third-party integration.

Vector (by Datadog)

Rust-based, high throughput, unified logs/metrics/traces, strong observability platform integration. Smaller memory footprint than Filebeat; comparable to Fluent Bit. Less battle-tested at Alibaba scale but broader community and enterprise support.

OpenTelemetry Collector

Vendor-neutral, multi-signal (traces, logs, metrics), extensive backend integrations, CNCF project. Heavier resource footprint than LoongCollector but mature ecosystem and no vendor lock-in to Alibaba Cloud SLS.

Software development agency

Build on loongcollector with DEV.co software developers

LoongCollector is best suited for teams processing high-volume observability data at scale, especially within Alibaba Cloud ecosystems. Pilot with one cluster and benchmark against Fluent Bit or Vector for your workload before full rollout. Contact us to assess fit with your architecture.

Talk to DEV.co

Related open-source tools

Surfaced by semantic similarity across the DEV.co open-source index.

Related on DEV.co

Explore the category and the services that help you build with it.

loongcollector FAQ

Does LoongCollector require Alibaba Cloud SLS to operate?
No, but remote configuration management via SLS console is a core feature. Standalone deployment possible via SDK or K8s Operator, though documentation suggests SLS integration is primary use case.
What performance gains should we realistically expect?
Benchmarks show 10x throughput (546 MB/s vs 36 MB/s) and 80% lower CPU/RAM at 10 MB/s processing load vs Fluent Bit. However, benchmarks are Alibaba-published; independent verification recommended before ROI calculations.
Is LoongCollector suitable for edge or resource-constrained deployments?
Partial. While designed for efficiency, C++ build complexity, kernel dependencies (eBPF), and 100+ plugin architecture add overhead. Fluent Bit or pure Golang collectors may be simpler for truly minimal deployments.
How stable is the project for production use?
Alibaba's 15+ year internal production history suggests stability, but open-source maturity is 2.5 years. 149 open issues indicate active development; pin versions and test breaking changes before upgrading.

Software development & web development with DEV.co

From first prototype to production, DEV.co delivers software development services around tools like loongcollector. Our software development agency staffs experienced software developers and web developers for custom software development, web development, integrations, and ongoing support across open-source observability and beyond.

Ready to evaluate LoongCollector for your infrastructure?

LoongCollector is best suited for teams processing high-volume observability data at scale, especially within Alibaba Cloud ecosystems. Pilot with one cluster and benchmark against Fluent Bit or Vector for your workload before full rollout. Contact us to assess fit with your architecture.