Leitmotif: A novel abstraction

Distributed tracing is gaining popularity and adoption, yet analyzing the end-to-end traces remains a daunting task. Traces are often large, containing hundreds of thousands of events, making them difficult to analyze and visualize. In this work, we aim to do the heavy lifting for analying distributed traces to debug performance problems by using frequent subgraph mining to identify frequent execution patterns in distirbuted systems. With these freqeunt patterns, called workflow motifs, we can guide users to which parts of their system should be optimized to improve the overall performance.