Category: service

ARMS Trace & Metrics Query

Query distributed traces and application performance metrics via ARMS Python SDK.

Prerequisites

• Install SDK (virtual environment recommended):

python3 -m venv .venv
. .venv/bin/activate
python -m pip install -U alibabacloud_arms20190808 alibabacloud_tea_openapi

• Configure environment variables:
  • ALIBABACLOUD_ACCESS_KEY_ID
  • ALIBABACLOUD_ACCESS_KEY_SECRET
  • ARMS_REGION_ID (e.g. cn-hangzhou)

Core Concepts

• Trace query: Search traces by service name, span name, duration, tags, error status. Time range in epoch milliseconds. Supports PID and exclusion filters.
• Trace detail: Retrieve full span tree for a single TraceID. Supports pagination for large traces.
• Method stack: Retrieve the method-level call stack for a specific span (requires TraceID, RpcID, and PID).
• Metrics query: Query predefined metric tables (e.g. appstat.incall) with dimensions, measures, and filters. Requires application PID. Supports ordering.
• Application PID: Unique app identifier in format xxx@regionId, obtained via ListTraceApps.

Quickstart — Search Traces (Python SDK)

import os, time
from alibabacloud_arms20190808.client import Client
from alibabacloud_arms20190808 import models
from alibabacloud_tea_openapi.models import Config

config = Config(
    access_key_id=os.environ["ALIBABACLOUD_ACCESS_KEY_ID"],
    access_key_secret=os.environ["ALIBABACLOUD_ACCESS_KEY_SECRET"],
    region_id=os.environ.get("ARMS_REGION_ID", "cn-hangzhou"),
)
config.endpoint = f'arms.{config.region_id}.aliyuncs.com'
client = Client(config)

end_ms = int(time.time() * 1000)
start_ms = end_ms - 15 * 60 * 1000

request = models.SearchTracesByPageRequest(
    region_id=config.region_id,
    start_time=start_ms,
    end_time=end_ms,
    service_name="my-app",
    min_duration=500,
    page_number=1,
    page_size=20,
)
response = client.search_traces_by_page(request)
for trace in response.body.page_bean.trace_infos:
    print(f"{trace.trace_id}  {trace.duration}ms  {trace.service_name}")

Quickstart — Query Metrics (Python SDK)

request = models.QueryMetricByPageRequest(
    region_id=config.region_id,
    metric="appstat.incall",
    start_time=start_ms,
    end_time=end_ms,
    interval_in_sec=60000,
    measures=["rt", "count", "error"],
    dimensions=["rpc"],
    filters=[models.QueryMetricByPageRequestFilters(key="pid", value="xxx@cn-hangzhou")],
    current_page=1,
    page_size=10,
)
response = client.query_metric_by_page(request)
for item in response.body.data.items:
    print(item)

Script Catalog

Search traces

python skills/observability/arms/aliyun-arms-query/scripts/search_traces.py \
  --service-name my-app \
  --last-minutes 15

Optional args: --region, --operation-name, --min-duration, --service-ip, --pid, --error-only, --tag KEY=VALUE (repeatable), --exclusion-filter KEY=VALUE (repeatable), --page, --page-size, --reverse.

Get trace detail

python skills/observability/arms/aliyun-arms-query/scripts/get_trace_detail.py \
  --trace-id 1c34ffee16xxxxxxxx

Optional args: --region, --start, --end, --page, --page-size.

Query metrics

python skills/observability/arms/aliyun-arms-query/scripts/query_metrics.py \
  --metric appstat.incall \
  --pid "xxx@cn-hangzhou" \
  --last-minutes 30

Optional args: --region, --measures (comma-separated), --dimensions (comma-separated), --interval, --page, --page-size, --order-by, --order ASC|DESC.

List traced applications

python skills/observability/arms/aliyun-arms-query/scripts/list_trace_apps.py

Optional args: --region, --app-type TRACE|EBPF, --resource-group-id.

Get span method stack

python skills/observability/arms/aliyun-arms-query/scripts/get_stack.py \
  --trace-id 1c34ffee16xxxxxxxx \
  --rpc-id 0.1 \
  --pid "xxx@cn-hangzhou"

Optional args: --region, --span-id, --start, --end.

Common Metric Names

Common measures: rt (response time), count (request count), error (error count). Common dimensions: rpc (endpoint), type (call type), exceptionClass.

Workflow

1. List traced apps with list_trace_apps.py to obtain PID.
2. Search traces with time range, service name, and optional filters via search_traces.py.
3. Get detail for a specific trace using its TraceID via get_trace_detail.py.
4. Inspect method stack for a slow or errored span via get_stack.py (requires TraceID, RpcID, PID).
5. Query aggregated metrics using PID, metric name, and desired measures/dimensions via query_metrics.py.

AccessKey Priority

1. Environment variables: ALIBABACLOUD_ACCESS_KEY_ID / ALIBABACLOUD_ACCESS_KEY_SECRET
2. Shared credentials file: ~/.alibabacloud/credentials (profile default)

API Discovery

• Product code: ARMS
• API version: 2019-08-08
• Metadata: https://api.aliyun.com/meta/v1/products/ARMS/versions/2019-08-08/api-docs.json

Validation

mkdir -p output/aliyun-arms-query
for f in skills/observability/arms/aliyun-arms-query/scripts/*.py; do
  python3 -m py_compile "$f"
done
echo "py_compile_ok" > output/aliyun-arms-query/validate.txt

Pass criteria: command exits 0 and output/aliyun-arms-query/validate.txt is generated.

Output And Evidence

• Save artifacts, command outputs, and API response summaries under output/aliyun-arms-query/.
• Include key parameters (region, PID, TraceID, time range) in evidence files for reproducibility.

References

• API reference: references/api-reference.md
• Source list: references/sources.md