A key component within the Dashboard is the interactive chart. The chart visualizes the volume of the events included within the current timeframe and View. These are visualized through the Current View series (in green). Filtering events within the grid is automatically reflected in the chart to visualize the events currently viewed.
The Event Chart displays how certain metrics have been performing over time. By default, the X-axis is time and the Y-axis is number of occurrences. The Y-axis will automatically adjust to the unit of measurement for each metric. The initial chart shows you the trends for the total number of errors over the selected time span. You can use the time selector to choose a different date, or use the mouse to zoom-in on a specific point in time. The relationship between the two can provide the context for the errors you’re having.
Set chart timeframes:
You can further customize those by changing the timeframes, selected applications, grid columns and more and save them as new View as described below.
- Last Hour - shows all events occurring within the last hour timeframe. This view is useful for giving you a real-time view of the behavior in your application.
- Last Day - this View provides a bird's eye view of all errors occurring within your application(s) over the last 24 hour window. This is useful view to see any spikes or fluctuations in errors as a result of a new deployment or environment change/outage.
- New Today - this View enables you to see all new errors that were introduced into your application(s) over the last day. This is a great view to use following a deployment to be able to see exactly if/what new errors were introduced to triage and analyze the most critical ones.
- New This Week - this View enables you to see all new errors that were introduced into your application(s) over the last week. This is a great view to use following one or more deployments to be able to see exactly if/what new errors were introduced to triage and analyze the most critical ones.
Whenever you select events in the grid they will be automatically visualized in the chart as the Current Selection series (in purple). Comparing both series enables you to see the effect of any number of errors locations on the whole.
Each series can be Pinned and Split:
Pinning a series - the current selection and view series can be pinned to keep it a separate series on the chart, even if you change selection. For example you may want to filter the Events List to show only NullPointerExceptions. You can then Pin the current view and clear the filter to contract the volume of NullPointerException related to the overall number of errors within a specific server, application or timeframe.
You can use button on the top right of the chart to remove all previously pinned series.
Splitting a series - you can use the split function to break any composite chart series such as the current view, selection or even a an error location which is being called into from multiple entry points to show the top 5 contributors as series on the graph. This enables you immediately see which ones have the most impact and may need to be dealt with first.
Hiding a series can be easily done by clicking the series legend to toggle On/Off its visibility.
Splitting the current View displays the IllegalStateException (orange) as the key contributor
Hover over any point in the graph to open the snapshot in the ARC screen from that point of time, directly. This tool enables you to drill into the cause of sudden spike in errors within your application. In case of composite series (one that aggregates multiple error types), you can perform the equivalent of a Split operation to break the series into its discrete components If the series you are hovering over is a composite series (i.e. it aggregates more than one error type) the tool window will enable you to perform the equivalent of a Split operation to break the series into its discrete components which can then be used to open a snapshot directly for.
Open a snapshot directly from the graph