In current large-scale IT systems, troubleshooting has become more complicated due to the diversification in the causes of failures, which has increased operational costs. Thus, clarifying the troubleshooting process also becomes important, though it is also time-consuming. We propose a method of automatically extracting a workflow, a graph indicating a troubleshooting process, using multiple trouble tickets. Our method extracts an operator's actions from free-format texts and aligns relative sentences between multiple trouble tickets. Our method uses a stochastic model to detect a resolution, a frequent action pattern that helps us understand how to solve a problem. We validated our method using real trouble-ticket data captured from a real network operation and showed that it can extract a workflow to identify the cause of a failure.

Recent carrier-grade networks use many network elements (switches, routers) and servers for various network-based services (e.g., video on demand, online gaming) that demand higher quality and better reliability. Network log data generated from these elements, such as router syslogs, are rich sources for quickly detecting the signs of critical failures to maintain service quality. However, log data contain a large number of text messages written in an unstructured format and contain various types of network events (e.g., operator's login, link down); thus, genuinely important log messages for network operation are difficult to find automatically. We propose a proactive failure-detection system for large-scale networks. It automatically finds abnormal patterns of log messages from a massive amount of data without requiring previous knowledge of data formats used and can detect critical failures before they occur. To handle unstructured log messages, the system has an online log-template-extraction part for automatically extracting the format of a log message. After template extraction, the system associates critical failures with the log data that appeared before them on the basis of supervised machine learning. By associating each log message with a log template, we can characterize the generation patterns of log messages, such as burstiness, not just the keywords in log messages (e.g. ERROR, FAIL). We used real log data collected from a large production network to validate our system and evaluated the system in detecting signs of actual failures of network equipment through a case study.

Network equipment, such as routers, switches, and RADIUS servers, generate various log messages induced by network events such as hardware failures and protocol flaps. In large production networks, analyzing the log messages is crucial for diagnosing network anomalies; however, it has become challenging due to the following two reasons. First, the log messages are composed of unstructured text messages generated in accordance with vendor-specific rules. Second, network events that induce the log messages span several geographical locations, network layers, protocols, and services. We developed a method to tackle these obstacles consisting of two techniques: statistical template extraction (STE) and log tensor factorization (LTF). The former leverages a statistical clustering technique to automatically extract primary templates from unstructured log messages. The latter builds a statistical model that collects spatial-temporal patterns of log messages. Such spatial-temporal patterns provide useful insights into understanding the impact and patterns of hidden network events. We evaluate our techniques using a massive amount of network log messages collected from a large operating network and confirm that our model fits the data well. We also investigate several case studies that validate the usefulness of our method.