Running the model
In this section, we will perform a basic interpretation of the results yielded by the model upon execution. To do so, follow the next steps:
Go to the top menu bar of Service Operations and click on the Service overview icon. The icon won’t be displayed if the model has not been run before. If that is the case, go to the Service Operations' administration section > Maps > Available maps and click on the play button that corresponds to the model we configured in the last sections of the documentation (e-commerce).
Once the model finishes loading and all queries and calculations are finished, the resulting topology map will be displayed on the screen. Since the cardinality of the entity users is relatively high, the topology model created automatically by auto-discovering all entities available in the original datasets will be significant (thousands of elements). To ease the analysis of the data process, click on the tree depth tool and set it to '2'.
Modify the default layout options to the following:
Orientation: 'top to bottom'
Click on the '+' icons on the highlighted entities in the next screenshot to expand their related subtrees:
This view allows users to gain a first level of understanding on how the modeled service is behaving:
The overall status of the model is set as "normal". The reason for this is that there are no top level entities with an "abnormal" (i.e., warning or critical) status. Since there is only one top level entity (e-commerce status) and its status is "normal", the overall model status takes its status from that entity.
The overall e-commerce status is "normal" because its impacting subsystems or elements (business status, UX status and app status) also run in normal condition on their side. Since we set all these entities to take a warning or critical status if more than half of their corresponding children have such status, and that condition is not met in neither case, they keep that status themselves.
The third level of business status shows the fundamental metrics (KPIs) that entity uses to calculate its own status. Clicking on any of them shows the evolution of that particular KPI over the analysis window. In this case, since we set the data availability window to one day, that is the historic data to be displayed in the time evolution chart for any of them. This mechanism allows for a drill-down analysis on the fundamental metrics and their behavior over time.
Analysis of issues
Despite the entire e-commerce service appearing in good healthy from all operational, business and user experience points of view, we might want to have a closer look at those entities that report an abnormal status. Although there are multiple ways to do so, we will follow a process hereby to identify, isolate and look into the details of those entities. To achieve that, follow the next steps:
Go to the Max depth control and select the maximum level of visibility to expand the hierarchical tree entirely.
Hover on the STATUS: NORMAL element and click on the down-pointing arrow that shows up when the bar is fully displayed. As a result, the component expands showing a tiled view of each and every entity discovered in the model, together with its current status and value and historic and trend information.
Click now on the CRITICAL indicator to filter out all entities that currently have that status. The model is automatically updated to show only the entities matching that filtering option. The filtering bar is also displayed on the upper part of the topology map.
Navigate across all entities now represented in the map. Click on any of them to display the detailed view component which will show, as before, the status of the selected entity, its status evolution over time, the relationships it has with other entities in the model, etc. Additionally, right click on the entity node to display other options:
On the detailed view menu, and under "affected by" or "impacts", click on the name of any entity listed in there for a quick navigation to that one. You can add additional filtering criteria to the current selection ('status = critical') by clicking on the type, sub-type and metadata values (where available). The next screenshot shows the result of clicking on the Operational value for an entity sub-type. The map then gets updated showing those entities in critical status whose subtype equals that value. Note also that the list of tiles on the left-hand component is updated to reflect that combination of filtering criteria.
To conclude this analysis, right-click on any of the entities in the map and select send all to monitor. Service Operations will project all those entities in a dashboard for a more detailed, combined analysis: