direkt zum Inhalt springen

direkt zum Hauptnavigationsmenü

Sie sind hier

TU Berlin

Page Content

ZerOps - A Self-Healing Platform

Telecommunication service and network operators are confronted with rising expectations towards availability, performance, and guaranteed QoS. The complexity of modern IT infrastructures has increased to a point, where traditional IT administration procedures fail to holistically ensure the dependability of the systems.

At the same time, various approaches around artificial intelligence (AI) are currently revolutionizing domains like medicine, manufacturing, or autonomous driving. This strongly motivates the utilization of AI for the autonomous management of highly complex IT systems (AIOps).

Researchers and global companies recognized this potential and started to work on AIOps solutions. Since 2015, the CIT department joint forces with industrial partners (Deutsche Telecom and Huawei Technologies Co., Ltd) to establish a joint research lab, working on solutions for anomaly detection/classification, predictive fault tolerance and auto-remediation. Thereby, the Self-Healing Cloud Platform ZerOps was developed, which is up to the current point constantly adjusted and enhanced by members of the CIT group.

Architecture

Lupe

The vision for ZerOps is to provide a scalable platform for monitoring, hierarchical in-place data analytics, and predictive system remediation. The term in-place refers to the explicit design goal to analyze collected data directly at the data source through streaming-based machine learning (ML) algorithms. ZerOps can be integrated in existing cloud infrastructures with. The second major design goal of ZerOps is a modular and flexible data analysis pipeline that can be assembled from multiple interchangeable elements. This allows customization to different infrastructure use cases, but also supports easy-to-use experimentation with new algorithmic approaches for research purposes. Due to the decentralized deployment, the data analysis is co-located with regular system parts. Therefore, its resource usage has to be limited to a certain percentage of the available resources. Furthermore, ZerOps incorporates streaming analytics as well as event aggregations to determine anomaly root causes and perform further advanced anomaly situation analyses. By the integration of unsupervised anomaly detection, ZerOps is able to detect unknown problems as well as already known and learned anomalies. A decentralized ML model repository enables transfer learning to overcome cold-start problems for dynamic IT-infrastructure components. ZerOps also supports automatic hyperparameter selection of ML algorithms.

Related Publications

W

Silent Consensus: Probabilistic Packet Sampling for Lightweight Network Monitoring

Wallschläger, Marcel and Acker, Alexander and Kao, Odej

Computational Science and Its Applications – ICCSA 2019. Springer International Publishing, 241–256. 2019

Download Bibtex entry

Unsupervised Anomaly Alerting for IoT-Gateway Monitoring using Adaptive Thresholds and Half-Space Trees

Wetzig, René and Gulenko, Anton and Schmidt, Florian

2019 Sixth International Conference on Internet of Things: Systems, Management and Security (IOTSMS). IEEE, 161–168. 2019

Download Bibtex entry

Decentralized Federated Learning Preserves Model and Data Privacy

Thorsten Wittkopp and Alexander Acker

18th International Conference on Service-Oriented Computing, To appear. 2020

Download Bibtex entry

MicroRCA: Root Cause Localization of Performance Issues in Microservices

Wu, Li and Tordsson, Johan and Elmroth, Erik and Kao, Odej

NOMS 2020-2020 IEEE/IFIP Network Operations and Management Symposium, 1–9. 2020

Link to publication Download Bibtex entry

Performance Diagnosis in Cloud Microservices using Deep Learning

Wu, Li and Bogatinovski, Jasmin and Nedelkoski, Sasho and Tordsson, Johan and Kao, Odej

18th International Conference on Service-Oriented Computing, To appear. 2020

Link to publication Download Bibtex entry

{

Anomaly Detection and Classification using Distributed Tracing and Deep Learning

S. Nedelkoski and J. Cardoso and O. Kao

2019 19th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (CCGRID). IEEE/ACM, 241-250. 2019

Download Bibtex entry

Anomaly Detection from System Tracing Data Using Multimodal Deep Learning

S. Nedelkoski and J. Cardoso and O. Kao

2019 IEEE 12th International Conference on Cloud Computing (CLOUD), 179-186. 2019

Download Bibtex entry

Learning more expressive joint distributions in multimodal variational methods

S. Nedelkoski and M. Bogojevski and O. Kao

2020 International Conference on Machine Learning, Optimization, and Data Science, LOD 2020, To appear. 2020

Download Bibtex entry

Self-Supervised Log parsing

S. Nedelkoski and J. Bogatinovski and A. Acker and J. Cardoso and O. Kao

European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, ECML-PKDD 2020, To appear. 2020

Download Bibtex entry

Zusatzinformationen / Extras

Quick Access:

Schnellnavigation zur Seite über Nummerneingabe