# Identifying Hotspots Using the `Dashboard` - `Key Visualizer`
![`smart`](ObserveHotspot01_aa_001_20231128.png)
![`smart`](ObverseHotspotBrightness_aa_001_20231129.png)
![`smart`](MonitorHotspot01_aa_001_20231020.png)

------------------------------------------------------------------------------
A hotspot refers to a situation in which not all nodes in a TiKV cluster are active. This means that only a few nodes are actively working, while the rest are idle.

Before discussing the resolution for the hotspot performance issue, let's first understand how to identify the hotspot in your TiDB. Generally, this can be achieved by utilizing a `Key Visualizer` view.

The `Key Visualizer` is a feature available in the `Dashboard` that is integrated with the TiDB cluster and is running on `Placement Driver` by default. In the left navigation panel, it is referred to as the `Key Visualizer`, where users can select the time range and monitor various dimensions from `IO` perspective such as `Write (bytes)`, `Read (bytes)`, `Write (keys)`, `Read (keys)`, and `All`.

In the slide, a screenshot of a specific `Key Visualizer` page view is presented. The horizontal axis depicts time, while the vertical axis represents TiKV Regions. TiKV Region blocks belonging to the same database are clustered together, as well as those belonging to the same table and index. Different brightness colors indicate distinct `IO` load indicators for each TiKV Region at a specific moment.

You can identify those hotspot TiKV Regions with significantly higher read or write traffic compared to other regions. The TiKV node where this region resides is highly likely to be a hotspot TiKV node.

In the graph: 
- The portion pointed out by `1` corresponds to which databases, tables, or indexes the squares represent.
- The portion pointed out by `2` represents some bright (hot) TiKV Regions, indicating that these regions have a significantly higher IO load for reads and writes compared to other regions at that moment.