数据采集可用性检测
该采集器用于检测数据采集的可用性和延迟。它会自动生成探测数据（指标、日志、链路追踪），然后通过 DQL 查询来验证数据是否成功采集，并测量从数据发送到可查询的延迟时间。
前置条件¶
- 需要配置 DataWay，用于数据上报和 DQL 查询
- 如果配置了
result_workspace，需要确保该工作空间 URL 可访问
配置¶
进入 DataKit 安装目录下的 conf.d/samples 目录，复制 ingestion_canary.conf.sample 并命名为 ingestion_canary.conf。示例如下：
[[inputs.ingestion_canary]]
## Collect interval, default 10m
interval = "10m"
## Query timeout, should be less than interval, default 5m
query_timeout = "5m"
## Poll interval for DQL query, default 500ms
poll_interval = "500ms"
## Max retry count for query errors, default 10
error_retries = 10
## Result workspace URL to report metrics
# result_workspace = "https://openway.guance.com?token=xxx"
## Data categories to collect (metric, logging, tracing)
## If not specified, all categories will be collected
categories = ["metric", "logging", "tracing"]
## Logging configuration
[inputs.ingestion_canary.logging]
storage_index = "default"
## Enable election mode
election = true
## Extra tags
[inputs.ingestion_canary.tags]
# some_tag = "some_value"
# more_tag = "some_other_value"
配置好后，重启 DataKit 即可。
可通过 ConfigMap 方式注入采集器配置 或 配置 ENV_DATAKIT_INPUTS 开启采集器。
指标说明¶
该采集器会生成两类数据：
- 测试数据：用于测试数据采集可用性的探测数据点（指标、日志、链路追踪）
- 结果指标：测试结果的指标数据，包含延迟时间和测试状态
探测数据点不携带全局标签，只包含数据点自身的字段和标签，以及配置中指定的 tags。
测试数据¶
ingestion_canary (指标)¶
Synthetic metric points emitted by the collector per probe round to verify metric ingestion path.
| Tags & Fields | Description |
|---|---|
| test_type ( tag) |
Test type: collect (collector) or cmd (CLI tool) |
| round | Round number of the ingestion canary probe. Type: int | (gauge) Unit: count Tagged by: test_type |
ingestion_canary (日志)¶
Synthetic logging payload emitted by the canary to verify logging ingestion path.
| Tags & Fields | Description |
|---|---|
| test_type ( tag) |
Test type: collect (collector) or cmd (CLI tool) |
| message | Synthetic log message emitted by the ingestion canary probe. Type: string Unit: N/A Tagged by: test_type |
| round | Round number of the ingestion canary probe. Type: int Unit: count Tagged by: test_type |
ingestion_canary (链路追踪)¶
Synthetic tracing payload emitted by the canary to verify trace collection and forwarding.
| Tags & Fields | Description |
|---|---|
| service ( tag) |
Service name |
| source ( tag) |
Source name |
| span_type ( tag) |
Span type |
| test_type ( tag) |
Test type: collect (collector) or cmd (CLI tool) |
| duration | Trace span duration in microseconds. Type: int Unit: time,μs Tagged by: service, source, span_type, test_type |
| parent_id | Parent span identifier emitted by the ingestion canary probe. Type: string Unit: N/A Tagged by: service, source, span_type, test_type |
| resource | Synthetic resource name emitted by the ingestion canary probe. Type: string Unit: N/A Tagged by: service, source, span_type, test_type |
| round | Round number of the ingestion canary probe. Type: int Unit: count Tagged by: service, source, span_type, test_type |
| span_id | Span identifier emitted by the ingestion canary probe. Type: string Unit: N/A Tagged by: service, source, span_type, test_type |
| start | Trace span start time as a Unix timestamp in microseconds. Type: int Unit: timeStamp,usec Tagged by: service, source, span_type, test_type |
| status | Synthetic span status emitted by the ingestion canary probe. Type: string Unit: N/A Tagged by: service, source, span_type, test_type |
| trace_id | Trace identifier emitted by the ingestion canary probe. Type: string Unit: N/A Tagged by: service, source, span_type, test_type |
结果指标¶
ingestion_canary_result¶
Result metrics of each ingestion canary probe round, including category, status, storage destination and latency.
| Tags & Fields | Description |
|---|---|
| category ( tag) |
Data category: M (metric), L (logging), T (tracing) |
| status ( tag) |
Test status: ok, timeout, error |
| storage_index ( tag) |
Storage index for logging data (optional) |
| latency_ms | Latency from feeding a canary point until it becomes queryable, in milliseconds. Type: int | (gauge) Unit: time,ms Tagged by: category, status, storage_index, test_type |
命令行工具¶
除了采集器模式，还提供了命令行工具用于一次性测试：
# 使用默认配置
datakit tool --ingestion-canary
# 指定日志数据的 storage index
datakit tool --ingestion-canary --ingestion-canary-index my_index
参数说明：
--ingestion-canary: 启用 ingestion canary 测试工具--ingestion-canary-index: 指定日志数据的 storage index，默认为 "default"（仅对日志数据有效）
功能说明：
该工具会生成一轮探测数据（指标、日志、链路追踪），发送到 DataWay 后持续查询直到找到数据或用户中断，并输出各数据类型的延迟时间。工具会持续运行，每轮测试间隔 10 秒。