> ## Documentation Index
> Fetch the complete documentation index at: https://docs.dqlabs.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Observability Dashboard — Pipeline

> How to read the Pipeline tab of the Observability Analytics dashboard — success rate, duration, failure reasons, and the slowest pipelines.

<script type="application/ld+json">
  {`{
              "@context": "https://schema.org",
              "@type": "TechArticle",
              "headline": "Observability Dashboard — Pipeline",
              "description": "How to read the Pipeline tab of the Observability Analytics dashboard — success rate, duration, failure reasons, and the slowest pipelines.",
              "url": "https://docs.dqlabs.ai/architecture/observability-pipeline-dashboard",
              "publisher": {
                "@type": "Organization",
                "name": "DQLabs Inc",
                "logo": "https://media.brand.dev/332adc35-5bc4-4d2b-bf78-256aa4a5e414.svg"
              }
              }`}
</script>

The **Pipeline** tab of the Observability Dashboard tracks the health of the pipelines and models that move and transform data — how often runs succeed, how long they take, why they fail, and which ones are slowest. Navigate to **Analytics → Observability (Data) → Pipeline** to access it.

## AI Summary

Same AI Summary component as the other Analytics tabs. In validation it sometimes continued to show the Data tab's summary text immediately after switching to Pipeline — give it a moment to refresh, or rely on the tiles and widgets below if it looks out of date.

## KPI cards

| Card | What it shows | How it's calculated |
| :- | :- | :- |
| **Pipeline Success Rate** | % of pipeline runs in the selected date range that completed successfully | Successful runs ÷ total runs, for runs matching the current filters and date range; compared against an **80.00% prior 90d** baseline |
| **Avg. Duration** | Mean run duration across in-scope pipeline runs | Average of (run end − run start) across matching runs. The badge shows both a duration delta **and** an execution count (e.g. "4010 executions vs prior 90d") — the execution count is the run volume the average is computed over, not a separate metric |
| **Failed Runs** | Count of pipeline runs that failed in the selected date range | A straight count of failed runs matching filters and date range |

<Warning>
  All three tiles animate with a count-up effect and can take several seconds to settle. Read a tile only after two consecutive looks agree.
</Warning>

## Widgets

### Success rate trend

A line chart of pipeline success rate (0-100%) over the selected date range, time-bucketed. In validation, this trend swung widely over the 90-day window (from under 20% up to roughly 80%), making this one of the more volatile trend widgets on any Analytics dashboard — worth checking before assuming a flat or near-100% baseline for your own pipelines.

### Failure reasons

A horizontal bar chart ranking failure causes by count (e.g. "Other failure", "Execution error"). Use this to see at a glance which failure category dominates before drilling into individual runs.

### Pipeline run outcome mix

A stacked bar chart of run outcomes (**Success**, **Failed**, **Skipped**) per time bucket across the date range. Unlike the Success rate trend's line, this shows absolute run volume per bucket, not just a rate — useful for distinguishing "the success rate dropped because of a few extra failures" from "the success rate dropped because run volume spiked."

### Slowest pipelines and models

A table ranking pipelines and models by average duration. Columns: **Pipeline / Model**, **Duration Trend** (a sparkline plus the delta vs. the start of the sparkline window), **Avg. Duration**, **Status** (**Healthy** or **Breaching** — whether the pipeline is currently within its configured SLA duration threshold).

## Filters

Same filter bar as the other Analytics dashboards: **Domain, Application, Product, Tag, Source, Asset Type, Asset, Date Range, Slice By, Previous Period**. See [Catalog Dashboard](/architecture/catalog-dashboard#filters) for the full list of options and the general filtering mechanism, which is identical here — applying a filter re-scopes every tile and widget to only the matching pipelines/runs.

<Tip>
  Filter selections persist across page reloads and across navigating away and back, for as long as your session is active. Click **Reset** to confirm you're looking at the true unfiltered baseline before drawing conclusions from the numbers.
</Tip>

## Slice By

**Slice By** (Day, Week, Month, Quarter, Half-year, Year; defaults to **Week**) rebuckets the **Success rate trend** and **Pipeline run outcome mix** charts' time axis, the same way it rebuckets the Quality Dashboard's trend chart and the Observability Data tab's heatmaps.

<Note>
  This dashboard's Slice-By behavior was not independently re-verified click-by-click during this validation pass — it follows directly from the pattern confirmed on three other Analytics dashboards (Quality, Catalog, Observability → Data): Slice By only rebuckets time-series visualizations and never changes a KPI tile. If you see different behavior in your environment, treat the live dashboard as the source of truth and let the docs team know via the feedback link so this page can be corrected.
</Note>

## Related pages

<CardGroup cols={2}>
  <Card title="Observability Dashboard — Data" icon="database" href="/architecture/observability-data-dashboard">
    Freshness, volume, and schema reliability for assets — the data-side counterpart to this pipeline-side dashboard
  </Card>

  <Card title="Observability Dashboard — Report" icon="file-chart-column" href="/architecture/observability-report-dashboard">
    Report and dashboard health, staleness, and usage
  </Card>
</CardGroup>


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.