Realizing the vision of using AI agents to automate critical IT tasks depends on the ability to measure and understand effectiveness of proposed solutions. We introduce ITBench, a framework that offers a systematic methodology for benchmarking AI agents to address real-world IT automation tasks. Our initial release targets three key areas: Site Reliability Engineering (SRE), Compliance and Security Operations (CISO), and Financial Operations (FinOps). The design enables AI researchers to understand the challenges and opportunities of AI agents for IT automation with push-button workflows and interpretable metrics. IT-Bench includes an initial set of 102 real-world scenarios, which can be easily extended by community contributions. Our results show that agents powered by state-of-the-art models resolve only 11.4% of SRE scenarios, 25.2% of CISO scenarios, and 25.8% of FinOps scenarios (excluding anomaly detection). For FinOps-specific anomaly detection (AD) scenarios, AI agents achieve an F1 score of 0.35. We expect ITBench to be a key enabler of AI-driven IT automation that is correct, safe, and fast. IT-Bench, along with a leaderboard and sample agent implementations, is available at https://github.com/ibm/itbench.

Misconfigurations are major causes of software failures. Existing practices rely on developer-written rules or test cases to validate configuration values, which are expensive. Machine learning (ML) for configuration validation is considered a promising direction, but has been facing challenges such as the need of large-scale field data and system-specific models. Recent advances in Large Language Models (LLMs) show promise in addressing some of the long-lasting limitations of ML-based configuration validation. We present the first analysis on the feasibility and effectiveness of using LLMs for configuration validation. We empirically evaluate LLMs as configuration validators by developing a generic LLM-based configuration validation framework, named Ciri. Ciri employs effective prompt engineering with few-shot learning based on both valid configuration and misconfiguration data. Ciri checks outputs from LLMs when producing results, addressing hallucination and nondeterminism of LLMs. We evaluate Ciri's validation effectiveness on eight popular LLMs using configuration data of ten widely deployed open-source systems. Our analysis (1) confirms the potential of using LLMs for configuration validation, (2) explores design space of LLM-based validators like Ciri, and (3) reveals open challenges such as ineffectiveness in detecting certain types of misconfigurations and biases towards popular configuration parameters.

Modern software projects have been increasingly using cloud services as important components. The cloud-based programming practice greatly simplifies software development by harvesting cloud benefits (e.g., high availability and elasticity). However, it imposes new challenges for software testing and analysis, due to opaqueness of cloud backends and monetary cost of invoking cloud services for continuous integration and deployment. As a result, cloud emulators are developed for offline development and testing, before online testing and deployment. This paper presents a systematic analysis of cloud emulators from the perspective of cloud-based software testing. Our goal is to (1) understand the discrepancies introduced by cloud emulation with regard to software quality assurance and deployment safety and (2) address inevitable gaps between emulated and real cloud services. The analysis results are concerning. Among 255 APIs of five cloud services from Azure and Amazon Web Services (AWS), we detected discrepant behavior between the emulated and real services in 94 (37 %) of the APIs. These discrepancies lead to inconsistent testing results, threatening deployment safety, introducing false alarms, and creating debuggability issues. The root causes are diverse, including accidental implementation defects and essential emulation challenges. We discuss potential solutions and develop a practical mitigation technique to address discrepancies of cloud emulators for software testing.

The ever-increasing demand for generative artificial intelligence (GenAI) has motivated cloud-based GenAI services such as Azure OpenAI Service. Like any large-scale cloud service, failures are inevitable in cloud-based GenAI services, resulting in user dissatisfaction and significant monetary losses. However, GenAI cloud services, featured by their massive parameter scales, hardware demands, and usage patterns, present unique challenges, including generated content quality issues and privacy concerns, compared to traditional cloud services. To understand the production reliability of GenAI cloud services, we analyzed production incidents from Microsoft spanning in the past four years. Our study (1) presents the general characteristics of GenAI cloud service incidents at different stages of the incident life cycle; (2) identifies the symptoms and impacts of these incidents on GenAI cloud service quality and availability; (3) uncovers why these incidents occurred and how they were resolved; (4) discusses open research challenges in terms of incident detection, triage, and mitigation, and sheds light on potential solutions.

In cloud-scale systems, failures are the norm. A distributed computing cluster exhibits hundreds of machine failures and thousands of disk failures; software bugs and misconfigurations are reported to be more frequent. The demand for autonomous, AI-driven reliability engineering continues to grow, as existing human-in-the-loop practices can hardly keep up with the scale of modern clouds. This paper presents STRATUS, an LLM-based multi-agent system for realizing autonomous Site Reliability Engineering (SRE) of cloud services. STRATUS consists of multiple specialized agents (e.g., for failure detection, diagnosis, mitigation), organized in a state machine to assist system-level safety reasoning and enforcement. We formalize a key safety specification of agentic SRE systems like STRATUS, termed Transactional No-Regression (TNR), which enables safe exploration and iteration. We show that TNR can effectively improve autonomous failure mitigation. STRATUS significantly outperforms state-of-the-art SRE agents in terms of success rate of failure mitigation problems in AIOpsLab and ITBench (two SRE benchmark suites), by at least 1.5 times across various models. STRATUS shows a promising path toward practical deployment of agentic systems for cloud reliability.

The rapid growth in the use of Large Language Models (LLMs) and AI Agents as part of software development and deployment is revolutionizing the information technology landscape. While code generation receives significant attention, a higher-impact application lies in using agents for the operational resilience of cloud services, which currently require significant human effort and domain knowledge. There is a growing interest in AI for IT Operations (AIOps), which aims to automate complex operational tasks, like fault localization and root cause analysis, reducing human intervention and customer impact. However, achieving the vision of autonomous and self-healing clouds through AIOps is hampered by the lack of standardized frameworks for building, evaluating, and improving AIOps agents. This vision paper lays the groundwork for such a framework by framing the requirements and then discussing design decisions that satisfy them. We also propose AIOpsLab, a prototype implementation leveraging agent-cloud-interface that orchestrates an application, injects real-time faults using chaos engineering, and interfaces with an agent to localize and resolve the faults. We report promising results and lay the groundwork to build a modular and robust framework for building, evaluating, and improving agents for autonomous clouds.

Ensuring the reliability and availability of cloud services necessitates efficient root cause analysis (RCA) for cloud incidents. Traditional RCA methods, which rely on manual investigations of data sources such as logs and traces, are often laborious, error-prone, and challenging for on-call engineers. In this paper, we introduce RCACopilot, an innovative on-call system empowered by the large language model for automating RCA of cloud incidents. RCACopilot matches incoming incidents to corresponding incident handlers based on their alert types, aggregates the critical runtime diagnostic information, predicts the incident's root cause category, and provides an explanatory narrative. We evaluate RCACopilot using a real-world dataset consisting of a year's worth of incidents from Microsoft. Our evaluation demonstrates that RCACopilot achieves RCA accuracy up to 0.766. Furthermore, the diagnostic information collection component of RCACopilot has been successfully in use at Microsoft for over four years.

Auditing, a central pillar of operating system security, has only recently come into its own as an active area of public research. This resurgent interest is due in large part to the notion of data provenance, a technique that iteratively parses audit log entries into a dependency graph that explains the history of system execution. Provenance facilitates precise threat detection and investigation through causal analysis of sophisticated intrusion behaviors. However, the absence of a foundational audit literature, combined with the rapid publication of recent findings, makes it difficult to gain a holistic picture of advancements and open challenges in the area.In this work, we survey and categorize the provenance-based system auditing literature, distilling contributions into a layered taxonomy based on the audit log capture and analysis pipeline. Recognizing that the Reduction Layer remains a key obstacle to the further proliferation of causal analysis technologies, we delve further on this issue by conducting an ambitious independent evaluation of 8 exemplar reduction techniques against the recently-released DARPA Transparent Computing datasets. Our experiments uncover that past approaches frequently prune an overlapping set of activities from audit logs, reducing the synergistic benefits from applying them in tandem; further, we observe an inverse relation between storage efficiency and anomaly detection performance. However, we also observe that log reduction techniques are able to synergize effectively with data compression, potentially reducing log retention costs by multiple orders of magnitude. We conclude by discussing promising future directions for the field.

System auditing provides a low-level view into cyber threats by monitoring system entity interactions. In response to advanced cyber-attacks, one prevalent solution is to apply data provenance analysis on audit records to search for anomalies (anomalous behaviors) or specifications of known attacks. However, existing approaches suffer from several limitations: 1) generating high volumes of false alarms, 2) relying on expert knowledge, or 3) producing coarse-grained detection signals. In this paper, we recognize the structural similarity between threat detection in cybersecurity and recommendation in information retrieval. By mapping security concepts of system entity interactions to recommendation concepts of user-item interactions, we identify cyber threats by predicting the preferences of a system entity on its interactive entities. Furthermore, inspired by the recent advances in modeling high-order connectivity via item side information in the recommendation, we transfer the insight to cyber threat analysis and customize an automated detection system, SHADEWATCHER. It fulfills the potential of high-order information in audit records via graph neural networks to improve detection effectiveness. Besides, we equip SHADEWATCHER with dynamic updates towards better generalization to false alarms. In our evaluation against both real-life and simulated cyber-attack scenarios, SHADEWATCHER shows its advantage in identifying threats with high precision and recall rates. Moreover, SHADEWATCHER is capable of pinpointing threats from nearly a million system entity interactions within seconds.