AIOps on Edge Computing Environments
The number of internet-connected devices and the amount of mobile traffic and internet traffic, in general, is rapidly increasing. The global research company Gartner predicts the amount of enterprise-generated data that was created and processed outside of traditional data centres to reach 75% in 2025. Together with upcoming network technologies like 5G, new application scenarios like autonomous vehicles, smart cities, sensor networks, industrial automation, or critical services and infrastructure control emerge.
Those application scenarios impose increased demands towards the
infrastructure, i. e.:
Support for a large number of devices
Traditional centralized cloud computing architectures cannot comply with these requirements anymore, due to the amount of data and congestion on the network layer. Also, in many situations, a response within predefined latency limits is not a soft requirement anymore but can have a serious impact on the system functionality in our everyday life.
Edge and Fog Computing describes a paradigm where Cloud Computing capabilities like compute, storage and network functions are provided close to the users as well as on the layer between public, private or hybrid clouds and the users. This results in a highly-distributed environment which not only constitutes an increase in complexity through the number of devices but further introduces new operational challenges: The tasks of operators are not only related to administering the data-centre cloud infrastructure but expand to the management of potentially thousands of remote edge sites and the maintenance of heterogeneous networks in-between.
In addition, edge devices are typically located outside of data centres and thus especially vulnerable to theft or damage, bad weather conditions, or network outages. Standard data centre procedures related to redundancy, access control, or maintenance need to be evaluated for their feasibility and adjusted to work in such environments. The challenges of Edge and Fog Computing environments create a paradox situation: a vulnerable infrastructure has a decisive impact on our everyday life, as it delivers crucial data for i.e. autonomous driving, connected healthcare or other critical processes. Managing this complexity to oversee the entire system and react with short intervals to comply with the high requirements on Edge Computing surpasses the ability of human experts. Concerning this, a new concept of combining AI methods to operate such complex infrastructures (AIOps) is on the rise. Starting as support systems for administrators, it proposes the vision of establishing a system able to autonomously operate and remediate large environments. In this project, we are applying AIOps methods to gradually automate administrative processes and to increase the availability, resilience, and fault-tolerance of Edge Computing environments.