Add Graph Insights and Watchdog Explains docs (#24141)

* Add Graph Insights and Watchdog Explains docs

* Fix title

* Fix reference link

* Add info div to graph insight for individual feature

* Add filter features and images

* Add suggestions from PM review

* Update images

* Fix links

* Apply suggestions from code review

* Fix broken link

* Apply suggestions from code review

Co-authored-by: Jen Gilbert <[email protected]>

---------

Co-authored-by: Jen Gilbert <[email protected]>

config/_default/menus/main.en.yaml

@@ -1286,11 +1286,21 @@ menu:
       url: dashboards/functions/beta/
       parent: dashboards_functions
       weight: 512
-    - name: Correlations
-      identifier: dashboards_correlations
-      url: dashboards/correlations/
+    - name: Graph Insights
+      identifier: dashboards_graph_insights
+      url: dashboards/graph_insights
       parent: dashboards
       weight: 6
+    - name: Metric Correlations
+      identifier: dashboards_correlations
+      url: dashboards/graph_insights/correlations/
+      parent: dashboards_graph_insights
+      weight: 601
+    - name: Watchdog Explains
+      identifier: dashboards_watchdog_explains
+      url: dashboards/graph_insights/watchdog_explains/
+      parent: dashboards_graph_insights
+      weight: 602
     - name: Template Variables
       url: dashboards/template_variables/
       parent: dashboards

content/en/dashboards/graph_insights/_index.md

@@ -0,0 +1,37 @@
+---
+title: Graph Insights
+disable_toc: false
+further_reading:
+- link: "/watchdog/insights/"
+  tag: "Documentation"
+  text: "Learn more about Watchdog Insights"
+---
+
+## Overview
+
+Graph insights can help you find potential root causes for an observed issue by searching for other metrics that exhibited irregular behavior around the same time. Metric Correlations scans your metrics from different sources, such as dashboards, integrations, APM, and custom metrics.
+
+## Metric Correlations
+
+<div class="alert alert-info">Metric Correlations is available for <a href="https://docs.datadoghq.com/dashboards/widgets/timeseries/">Timeseries widgets</a> with the <strong>Metric</strong> data source.</div>
+
+To target the search more effectively, Metric Correlations uses information about related dashboards and services. Correlations can sift through metrics from various sources, including APM, integrations, and dashboards, as well as arbitrary metric namespaces you select. It searches for irregularities in other metrics over the corresponding time period, enabling Datadog to automatically provide clues that facilitate a more efficient root cause analysis.
+
+For more information, see the [Metric Correlations][1] documentation.
+
+## Watchdog Explains
+
+<div class="alert alert-info">Watchdog Explains is available for <a href="https://docs.datadoghq.com/dashboards/widgets/timeseries/">Timeseries widgets</a> with the <strong>Metric</strong> data source.</div>
+
+Datadog collects various types of data to provide insights into application performance, including metrics, traces, and logs, which tell you what, how, and why something is happening. Watchdog Explains analyzes high-level trends such as latency, error rates, or request count evolution to detect critical signals. Upon observing a spike in these graphs, Watchdog Explains helps you investigate the immediate questions:
+- What is the source of the spike?
+- Does this anomaly affect everyone or is an isolated incident?
+
+For more information, see the [Watchdog Explains][2] documentation.
+
+## Further reading
+
+{{< partial name="whats-next/whats-next.html" >}}
+
+[1]: /dashboards/graph_insights/correlations/
+[2]: /dashboards/graph_insights/watchdog_explains/

content/en/dashboards/correlations/_index.md → content/en/dashboards/graph_insights/correlations.md

@@ -2,6 +2,7 @@
 title: Metric Correlations
 aliases:
     - /graphing/correlations/
+    - /dashboards/correlations/
 further_reading:
 - link: "/dashboards/"
   tag: "Documentation"
@@ -23,7 +24,7 @@ further_reading:
 
 Metric Correlations can help you find potential root causes for an observed issue by searching for other metrics that exhibited irregular behavior around the same time. Correlations scans your metrics from different sources such as dashboards, integrations, APM, and custom metrics.
 
-## Search
+## Find correlated metrics
 
 You can start your metric correlations exploration from any of your dashboards, notebooks, APM, Watchdog alerts, or monitor status pages.
 
@@ -38,17 +39,15 @@ Correlations *tries* to automatically detect the area of interest (anomalous beh
 
 **Note**: Correlation searches are available for a single metric. For graphs with multiple metrics, select the series of interest. From a full-screen graph, select one series on the graph legend, then click the **Correlations** tab.
 
-### Edit
+## Customize your search
 
 You can customize the default search parameters of correlations. From a full-screen graph, on the *Correlations* tab, click the **Edit Search** button, or click directly on the graph.
 
-* Click and drag on the graph to set the timeframe for your correlations search. If an area is already selected (pink box), you can move or resize the selection.
-* Define the sources you want correlations to search from (APM services, integrations, dashboards, or custom metrics).
-* `Auto-select` or `Custom select` from specific categories. For custom metrics, at least one selection is required.
-* Custom metrics is the only category not selected by default. Choose metric namespaces or single metrics to search correlations upon.
-* Use the tag filter box to scope the search by a tag.
-
-### Results
+1. Click and drag on the graph to set the time frame for your correlations search. 
+1. Define the sources you want correlations to search from (APM services, integrations, dashboards, or custom metrics).
+1. Select **Auto-select** or **Custom select** from specific categories. For custom metrics, at least one selection is required.
+   * Custom metrics is the only category not selected by default. Choose metric namespaces or single metrics to search correlations upon.
+1. Use the tag filter box to scope the search by a tag.
 
 A list of search results is displayed below the search graph with the following:
 
@@ -59,7 +58,7 @@ A list of search results is displayed below the search graph with the following:
 
 {{< img src="dashboards/correlations/search_results.png" alt="Search results" style="width:80%;">}}
 
-As results load, you can explore the details without waiting for all results. When the search is finished, the message "Search completed!" appears.
+As results load, you can explore the details without waiting for all results. 
 
 ## Investigate
 

content/en/dashboards/graph_insights/watchdog_explains.md

@@ -0,0 +1,48 @@
+---
+title: Watchdog Explains
+aliases:
+    - /graphing/correlations/
+    - /dashboards/correlations/
+further_reading:
+- link: "/watchdog/insights/"
+  tag: "Documentation"
+  text: "Learn more about Watchdog Insights"
+---
+
+## Overview
+
+{{< beta-callout url="https://www.datadoghq.com/private-beta/watchdog-explains-graph-insights/">}}
+Watchdog Explains is available in Private Beta. To request access, complete the form.
+{{< /beta-callout >}}
+
+<div class="alert alert-info">Watchdog Explains is available for <a href="https://docs.datadoghq.com/dashboards/widgets/timeseries/">Timeseries widgets</a> with the <strong>Metric</strong> data source.</div>
+
+{{< img src="dashboards/graph_insights/watchdog_explains/watchdog_explains_walkthrough.mp4" alt="A walkthrough of the Watchdog Explains product" video=true >}}
+
+Watchdog Explains is an investigation assitant that guides you to the root cause of anomalies on any timeseries graph. 
+
+In Datadog, an investigation typically starts with graphs, then branches out into investigating individual assets. Watchdog Explains makes investigations more efficient by automatically showing which individual tags account could be responsible for a given spike. This allows you to focus your investigation on problematic areas of your infrastructure or software stack.
+
+## How does it work?
+
+1. **Watchdog Explains runs anomaly detection** and determines if the graph shape or value changed from the historical pattern. It scans metric-based graphs to look for anomalies and dissects the anomaly to show which tags are responsible. 
+
+2. **Then, it runs the same query filtered on each applicable tag group**. It compares the same timeseries data across each applicable tag group against the source graph to identify which ones represent that anomalous behavior. 
+   - If a graph’s shape changes significantly by removing an individual tag group, it infers that the tag is mostly likely the cause of the spike. 
+   - Watchdog Explains shows you evidence to quantify exactly how influential a given tag is.
+
+
+## Investigate anomalies
+
+Start your investigation from any timeseries metric graph. Open a graph in full screen to trigger Watchdog Explains.
+
+{{< img src="dashboards/graph_insights/watchdog_explains/graph_anomaly_detection.png" alt="Watchdog Explains highlights the anomalous parts of a graph based on historical data" style="width:90%;" >}}
+
+Watchdog Explains highlights anomalies with a pink box. On the right side panel, you can view the tags that are responsible for the spike. Click on a tag to see evidence of how it contributes to the graph shape.
+
+{{< img src="dashboards/graph_insights/watchdog_explains/graph_filter_tag.png" alt="Filter out the offending tag, in this case researcher-query, to compare the original against what the graph would look like without the offending tag" style="width:90%;" >}}
+
+
+## Further reading
+
+{{< partial name="whats-next/whats-next.html" >}}

content/en/watchdog/insights.md

@@ -68,6 +68,14 @@ To share a given outlier, click on it in the insight panel to open the details s
 
 The link to the outlier expires with the retention of the underlying data. For instance, if the logs used to build the outlier are retained for 15 days, the link to the outlier expires with the logs after 15 days.
 
+## Explore graph insights with Watchdog explains
+{{< img src="dashboards/graph_insights/watchdog_explains/graph_filter_tag.png" alt="Filter out the offending tag, in this case researcher-query, to compare the original against what the graph would look like without the offending tag" style="width:90%;" >}}
+Datadog collects various types of data to provide insights into application performance, including metrics, traces, and logs, which tell you what, how, and why something is happening. Watchdog Explains analyzes high-level trends such as latency, error rates, or request count evolution to detect critical signals. Upon observing a spike in these graphs, Watchdog Explains helps you investigate the immediate questions:
+- What is the source of the spike?
+- Does this anomaly affect everyone or is an isolated incident?
+
+For more information, see the [Watchdog Explains][14] documentation.
+
 ## Outlier types
 
 {{< tabs >}}
@@ -296,3 +304,4 @@ For Kubernetes Explorer, the Watchdog Insight carousel reflects [all the Kuberne
 [11]: /watchdog/#overview
 [12]: https://app.datadoghq.com/watchdog
 [13]: https://app.datadoghq.com/rum/error-tracking
+[14]: /dashboards/graph_insights/watchdog_explains