COLD

Cold start

The job waits for its first executors before any task can run.

Outcome: Speed
Scope: Judged on the whole run
Usual fix: A change in the job code
CI gate: No budget flag: gate it with --fail-on-introduced against a baseline

An example finding

Where you see it

Dashboard

The COLD tag marks it in the action region of the dashboard, among the widgets that rank a fix. How to read it in the findings guide.

Agent (MCP)

An assistant can read this entry with the list_detectors tool and its background with get_finding_documentation. The call and the response below come from a real run of the MCP server.

Call

{
  "tool": "list_detectors",
  "arguments": {
    "type": "coldStart"
  }
}

Response

{
  "count": 1,
  "detectors": [
    {
      "type": "coldStart",
      "name": "cold start",
      "tag": "COLD",
      "tagSharedWith": [],
      "family": "speed",
      "pitch": "The job waits for its first executors before any task can run.",
      "region": "action",
      "scope": "app",
      "fixEffort": "code",
      "thresholdSummary": "the first stage waiting over 30s for an executor",
      "detectors": [
        {
          "detector": "coldStart",
          "version": 1,
          "thresholds": {
            "gapSeconds": 30
          }
        }
      ],
      "pages": {
        "catalog": "docs/detectors/coldStart.html",
        "guide": "docs/user-guide/understanding-findings.html#cold"
      },
      "tuning": {
        "kind": "page",
        "anchor": "#bottleneck-cold-start",
        "path": "docs/tuning-reference/bottleneck-cold-start.html#bottleneck-cold-start"
      },
      "gate": null
    }
  ]
}

get_finding_documentation call

{
  "tool": "get_finding_documentation",
  "arguments": {
    "type": "coldStart"
  }
}

Its response, with each Markdown body replaced by its length

{
  "type": "coldStart",
  "name": "Cold Start",
  "detectionDoc": {
    "tag": "COLD",
    "title": "Executor cold start",
    "content": "<456 characters of Markdown>"
  },
  "tuningDoc": {
    "anchor": "#bottleneck-cold-start",
    "title": "Cold Start",
    "content": "<4950 characters of Markdown>"
  }
}

CI (CLI)

List this entry from the command line. The output below is a real run of the command. Budgets and exit codes are in the CI guide.

sparkforensics-analyze detectors --type coldStart --format md

# Detectors (1)

## Speed

| Tag | Type | Scope | Fix | Gate | What it finds |
| --- | --- | --- | --- | --- | --- |
| COLD | `coldStart` | app | code |  | The job waits for its first executors before any task can run. |

Report only this finding type for a run with --type:

sparkforensics-analyze <event-log> --type coldStart --impact critical,warning,info

It has no budget flag. Fail a build on a new one by comparing against a baseline with --fail-on-introduced.

Airflow

The operator has no budget parameter for it. Use fail_on_introduced with a baseline.

check_spark_job = SparkForensicsOperator(
    task_id="check_spark_job",
    log_source=...,
    backend=SubprocessAnalyzeHook(),
    baseline_log_source=...,
    fail_on_introduced="critical",
)

The operator works per run, and its summary counts findings per impact band. See the operator for its parameters.

How it is measured

Thresholds

The first stage waiting over 30s for an executor.

coldStart version 1

Threshold	Default
`gapSeconds`	30

Each rule's impact bands and fixed fallbacks are in the threshold tables.

Override a default per run with --thresholds; see Tuning detector thresholds.

Docs

This finding has its own tuning page: Spark tuning reference.
Findings guide: COLD in Understanding findings.