Trace Heatmap and K-Lens
- Pralav Dessai
Kloudfuse 3.0 has two powerful APM observability tools - Trace Heatmap and K-Lens to assist in service-level issue detection and advanced debugging using APM Trace data.
Trace Heatmap
The Kloudfuse Trace Heatmap is a trace-data visualization tool that helps visually detect deviations and outliers in latencies reported by APM data. The Trace Heatmap is a grid of time-axis (horizontal) and duration-axis (vertical). Each cell of the Trace Heatmap captures the count of events (number of spans) that are detected within the time-bucket and duration-bucket. The visualization also uses deeper colors for higher span-counts, and lighter color for lower span-counts to provide a quick indication of the distribution of span-latencies over a selected time-period.
The Trace Heatmap also allows users to select a region on the Heatmap, with options to view spans/events or start advanced analysis using K-Lens.
K-Lens
K-Lens is a proprietary APM data analysis tool that provides clues into “what went wrong” by comparing thousands of attributes across span events. Users can select a region on the Trace Heatmap that they wish to compare against the baseline (all other events in the Heatmap). The K-Lens data analysis algorithm checks every attribute of the events in the Heatmap, and then figures out if there are major differences in the values for these attributes between the selection area and baseline. The resulting attribute charts are ranked and charts of higher significance are presented at the top.
Users can also add filters from the data presented in the charts, to allow in-depth investigations by querying the Heatmap again, and performing subsequent analysis using K-Lens on returned Heatmap.
