Hi! So, you've just installed StackState and you are ready to get started.
The first step to take is integrating StackState with your IT systems. This can be done by installing one or more StackPacks.
After installing one or more StackPacks, you can go to the Explore Mode to explore your IT landscape or visit a specific view from your installed StackPacks.
StackState visualizes components by the layer and domain they are placed in. These are logical groupings of components. Layers are displayed on the vertical axis. Domains are displayed on the horizontal axis.
A topology consists of components and relations combined with their health state. Because topologies can get very large, StackState automatically groups the components.
The health state of a component is indicated by two colors:
The inner color of a component indicates the health state that is calculated for the component itself.
The outer color indicates the potential impact of components that this component depends on.
The direction of a relation's arrow indicates dependency. For example,
app -> db means:
app depends on
db. Health propagates in the opposite direction. So if the
db component turns red, the
app component will turn red too.
The timeline at the bottom of the screen gives you the ability to go to any point in time. All the information that you see (component details, metric streams, etc.) is relative to the time that is currently selected. Normally, StackState is in the live mode, this means that StackState automatically displays the latest state of the stack.
To see the details of a component or a relation, you can click on it. Click on "Show component properties" to see all details of a component.
Both components and relations can have one or multiple telemetry streams. The most common type is a metric stream also known as time series. If you click on a metric stream, you can see the metric stream in a popup.
If you click on an event stream, you can see the event/log stream in a popup. Again, there are a number of drill-down capabilities available on the left side of the popup.
To quickly find the cause of any deviating component, head to the right-hand pane in the selected view where you can find the Problem clusters section of the sidebar. It reduces the cognitive flow and provides immediate understanding of ongoing problems. The Problem clusters pane will show you the cause of problems in your view. This summary is based on the components impacted in your current view combined with all (potential) causes. Problems are automatically clustered by their root cause.
Problems and issues are displayed in the following order order:
Problems are prioritized by the number of issues related to the specific problem. More affected components means higher priority on the list.
In case of multiple problems having the same number of issues, StackState prioritizes the most recent one and presents the rest from the newest on the top to the oldest on the bottom.
Component-specific issues grouped inside problems are displayed from the oldest ones at the top to the most recent issue at the bottom of the list.
Not all components displayed in the problem pane are necessarily also visible in the current topology view. To make the root cause of a problem visible, hover a mouse pointer over the component and select
show -> root cause from the
Quick Actions menu.
To show all events for the selected Topology, select the Events perspective from the top of the screen. Examples of important events that may appear here are health state changes and changes to the components themselves, like version changes. With event handlers, you can configure StackState to react to any events, for example, by automatically creating a ticket or triggering some automation.
The Events perspective is not the only place you can find events; you can find the latest 5 events in the Events section of the right-hand pane that presents information about the selected Topology.