aliyun-trace

What it is and what it does

Aliyun Trace is a Python library for creating manual traces of LLM applications using Aliyun's tracing infrastructure. It builds on OpenTelemetry standards (via opentelemetry-api and opentelemetry-semantic-conventions) to instrument LLM workloads, allowing developers to capture and export trace data to Aliyun's Application Real-time Monitoring Service (ARMS). The library wraps instrumentation logic using wrapt and applies Aliyun-specific semantic conventions.

The package is positioned as a bridge between your LLM code and Aliyun's observability platform. It handles the mechanics of trace creation so you can focus on marking the spans and events that matter in your application. With low install friction and support for Python 3.7 through 3.11, it integrates into existing OpenTelemetry setups, though its aging maintenance status suggests it may lag behind rapid changes in the OpenTelemetry ecosystem.

Use it for:

• Instrument LLM inference pipelines running on Aliyun infrastructure to capture request flow and latency.
• Create manual traces for custom LLM workflows where automatic instrumentation is insufficient.
• Export trace data to Aliyun ARMS for centralized monitoring and debugging of distributed LLM applications.
• Integrate LLM observability into existing OpenTelemetry-based monitoring stacks on Aliyun.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Manually creates traces for LLM applications using Aliyun's tracing infrastructure, integrating with OpenTelemetry for standardized observability.

Yes, if you are running LLM workloads on Aliyun and need to send traces to ARMS. The low install friction and permissive license make it straightforward to add. However, the 420-day maintenance gap warrants checking that it remains compatible with your OpenTelemetry version and Aliyun ARMS API before committing to production use.

Install

aliyun-trace on PyPI

pip install aliyun-trace

uv add aliyun-trace

poetry add aliyun-trace

Installing aliyun-trace

Before you install

Low install friction with a pure-Python wheel. Maintenance status is aging—last release was 420 days ago—so expect limited active development and potential staleness relative to OpenTelemetry ecosystem changes.

License in practice

Apache-2.0 permissive license allows commercial and private use with minimal restrictions, typical for observability libraries.

Quickstart

pip install aliyun-trace

from aliyun_trace import Tracer
tracer = Tracer()
tracer.create_trace()

Requires Python 3.7 or later; depends on aliyun-semantic-conventions, opentelemetry-api, opentelemetry-semantic-conventions, and wrapt being available.

Verify before relying

• Exact API surface and trace creation methods not detailed in excerpt; confirm whether it covers your LLM instrumentation needs.
• Whether this integrates with Aliyun ARMS console or requires additional configuration beyond the library itself.
• Current compatibility with recent OpenTelemetry releases, given 420-day gap since last update.

Package facts

Evidence: aliyun_trace-1.0.3-py3-none-any.whl

Tags

More Monitoring packages