PRAXIS: Integrating Program Analysis with Observability for Root-Cause Analysis

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Cui, Shengkun, Krishna, Rahul, Jha, Saurabh, Iyer, Ravishankar K.
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911629770227712
author Cui, Shengkun
Krishna, Rahul
Jha, Saurabh
Iyer, Ravishankar K.
author_facet Cui, Shengkun
Krishna, Rahul
Jha, Saurabh
Iyer, Ravishankar K.
contents Unresolved production cloud incidents cost an average of over $2M per hour. This paper introduces PRAXIS, an orchestrator that manages and deploys an agentic workflow for diagnosing code- and configuration-caused cloud incidents. PRAXIS employs an LLM-driven structured traversal over two types of graph: (1) a service dependency graph (SDG) that captures microservice-level dependencies; and (2) a hammock-block program dependence graph (PDG) that captures code-level dependencies for each microservice. Compared to state-of-the-art ReAct baselines, PRAXIS improves RCA accuracy by up to 6.3x while reducing token consumption by 5.3x. PRAXIS is demonstrated on a set of 30 comprehensive real-world incidents that is being compiled into an RCA benchmark.
format Preprint
id arxiv_https___arxiv_org_abs_2512_22113
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle PRAXIS: Integrating Program Analysis with Observability for Root-Cause Analysis
Cui, Shengkun
Krishna, Rahul
Jha, Saurabh
Iyer, Ravishankar K.
Distributed, Parallel, and Cluster Computing
Artificial Intelligence
Software Engineering
Unresolved production cloud incidents cost an average of over $2M per hour. This paper introduces PRAXIS, an orchestrator that manages and deploys an agentic workflow for diagnosing code- and configuration-caused cloud incidents. PRAXIS employs an LLM-driven structured traversal over two types of graph: (1) a service dependency graph (SDG) that captures microservice-level dependencies; and (2) a hammock-block program dependence graph (PDG) that captures code-level dependencies for each microservice. Compared to state-of-the-art ReAct baselines, PRAXIS improves RCA accuracy by up to 6.3x while reducing token consumption by 5.3x. PRAXIS is demonstrated on a set of 30 comprehensive real-world incidents that is being compiled into an RCA benchmark.
title PRAXIS: Integrating Program Analysis with Observability for Root-Cause Analysis
topic Distributed, Parallel, and Cluster Computing
Artificial Intelligence
Software Engineering
url https://arxiv.org/abs/2512.22113