Developers
August 10, 2020

Microsoft Azure Working With AI: AIOpS For DevOps

AIOpS, the infusion of AI into the cloud while using DevOps.

Today we will talk about AIOpS. What exactly is AIOpS? It's the infusion of AI into the cloud and the use of DevOps. This helps the platform become more adaptive and efficient.  

Building and operating a global cloud infrastructure at the scale Azure does, it's not an easy task. The components are always evolving and the system needs constant work to run successfully. Azure currently counts with over 160 data centers and it works in 60 regions.

To face the challenge, Microsoft has created a special team called the AIOpS that helps collaborate in Azure services. AIOpS works together with the Microsoft Research team and they develop aAI solutions for cloud management.  

Why should I use AIOpS?

To build a reliable cloud service one has to be able to manage data efficiently and correctly. In the Azure case, there are thousands of developers, operational engineers, customer support engineers, and program managers working on this so the cloud remains a stable and safe place.  

The cloud is always increasing in its scalability and complexity. This means that the systems are always needing more hands working on it as it grows. There is also a constant increase in customers and partners. This results in a steady increase in workloads.  

AIOpS is transforming the cloud business. How is it achieving this? It's offering improved service quality and customer experience. This results in improved performance and productivity of engineers and improved efficiency of the platform itself.

AIOpS counts with higher service quality and efficiency. The services come with the capability of self-monitoring, without the need for human intervention. Automation is powered and it improves service quality, reliability, availability, and performance.  

By using AI and ML engineers are free from investigating repeated processes and from manually operating the services. As Artificial Intelligence is set up to work by itself, the human resources are allocated to solving new problems and building new functions.  

AIOpS enables customers to use maintain and solve workloads on top of the provided cloud services in an easy way. It is recommended to use AIOpS to improve the knowledge of customer satisfaction better.  

How are AIOpS built?

The solution is built based on data. Measurements of systems, customers, and processes are used as key data for AIOpS solutions. Then the data is distilled into insights obtained from system behavior and customer behavior. 

The insights obtained from the data range from identifying problems, identifying the causes of the problems, and predicting what will happen in the future. The insights are always correlated to business metrics.  This way businesses can use the insights specifically to meet their business needs.

The use of AI reduces operational costs, improves the systems, and increases customer satisfaction overall. This doesn't only apply to AIOpS but to any service that runs powered by AI. This has been seen in the Azure platform, especially in the Azure IaaS virtual machines.

Engineering improvements and system innovation are primordial for the continuous development of the platform. For this, Azure provides Hardware Failure Prediction, Pre-Provisioning Service, Incident management, and Anomaly Detection 

Hardware Failure Prediction protects cloud customers from any interruption caused by hardware failures. The prediction has also been expanded to other types of hardware issues besides Azure, for example, memory and networking router failures.

The Pre-Provisioning Service works in Azure, and it brings Virtual Machine deployment reliability. It works by creating pre-provisioned Virtual Machines. These are configured ahead of customer requests. It leverages a prediction algorithm, predicting configurations and the number of VMs per configuration that have to be created. 

Incident management identifies and mitigates platform outages. Any incident management that occurs in the system, it is detected and mitigated by the Incident Management. There are three Key Performance Indicators (KPI) to measure how well the problems are solved. The time to detect (TTD), the time to engage (TTE), and the time to mitigate (TTM).

Anomaly Detection provides monitoring and anomaly detection for Azure IaaS. This detection solution works by targeting a broad spectrum of anomaly patterns, including generic patterns defined by thresholds. Other patters are included too, such as leaking patterns (such as memory leakages) and emerging patterns.  

In conclusion, AIOpS is the infusion of AI into the cloud. This helps the platform become more adaptive and efficient. The Azure platform is always evolving and the system always needs to be improved to be maintained and up to date. Azure is not a small cloud ecosystem, we are talking about 160 datacenters that work on  60 regions. Currently, AIOpS is working hand by hand with the Microsoft Research team. They develop AI solutions for the cloud. AIOpS is based on data. The data is then transformed and distilled into insights. The insights are the most useful data for businesses, as it is what makes them advance.

TagsDevOpsAICloudAIOpS
Lucas Bonder
Technical Writer
Lucas is an Entrepreneur, Web Developer, and Article Writer about Technology.

Related Articles

Back
DevelopersAugust 10, 2020
Microsoft Azure Working With AI: AIOpS For DevOps
AIOpS, the infusion of AI into the cloud while using DevOps.

Today we will talk about AIOpS. What exactly is AIOpS? It's the infusion of AI into the cloud and the use of DevOps. This helps the platform become more adaptive and efficient.  

Building and operating a global cloud infrastructure at the scale Azure does, it's not an easy task. The components are always evolving and the system needs constant work to run successfully. Azure currently counts with over 160 data centers and it works in 60 regions.

To face the challenge, Microsoft has created a special team called the AIOpS that helps collaborate in Azure services. AIOpS works together with the Microsoft Research team and they develop aAI solutions for cloud management.  

Why should I use AIOpS?

To build a reliable cloud service one has to be able to manage data efficiently and correctly. In the Azure case, there are thousands of developers, operational engineers, customer support engineers, and program managers working on this so the cloud remains a stable and safe place.  

The cloud is always increasing in its scalability and complexity. This means that the systems are always needing more hands working on it as it grows. There is also a constant increase in customers and partners. This results in a steady increase in workloads.  

AIOpS is transforming the cloud business. How is it achieving this? It's offering improved service quality and customer experience. This results in improved performance and productivity of engineers and improved efficiency of the platform itself.

AIOpS counts with higher service quality and efficiency. The services come with the capability of self-monitoring, without the need for human intervention. Automation is powered and it improves service quality, reliability, availability, and performance.  

By using AI and ML engineers are free from investigating repeated processes and from manually operating the services. As Artificial Intelligence is set up to work by itself, the human resources are allocated to solving new problems and building new functions.  

AIOpS enables customers to use maintain and solve workloads on top of the provided cloud services in an easy way. It is recommended to use AIOpS to improve the knowledge of customer satisfaction better.  

How are AIOpS built?

The solution is built based on data. Measurements of systems, customers, and processes are used as key data for AIOpS solutions. Then the data is distilled into insights obtained from system behavior and customer behavior. 

The insights obtained from the data range from identifying problems, identifying the causes of the problems, and predicting what will happen in the future. The insights are always correlated to business metrics.  This way businesses can use the insights specifically to meet their business needs.

The use of AI reduces operational costs, improves the systems, and increases customer satisfaction overall. This doesn't only apply to AIOpS but to any service that runs powered by AI. This has been seen in the Azure platform, especially in the Azure IaaS virtual machines.

Engineering improvements and system innovation are primordial for the continuous development of the platform. For this, Azure provides Hardware Failure Prediction, Pre-Provisioning Service, Incident management, and Anomaly Detection 

Hardware Failure Prediction protects cloud customers from any interruption caused by hardware failures. The prediction has also been expanded to other types of hardware issues besides Azure, for example, memory and networking router failures.

The Pre-Provisioning Service works in Azure, and it brings Virtual Machine deployment reliability. It works by creating pre-provisioned Virtual Machines. These are configured ahead of customer requests. It leverages a prediction algorithm, predicting configurations and the number of VMs per configuration that have to be created. 

Incident management identifies and mitigates platform outages. Any incident management that occurs in the system, it is detected and mitigated by the Incident Management. There are three Key Performance Indicators (KPI) to measure how well the problems are solved. The time to detect (TTD), the time to engage (TTE), and the time to mitigate (TTM).

Anomaly Detection provides monitoring and anomaly detection for Azure IaaS. This detection solution works by targeting a broad spectrum of anomaly patterns, including generic patterns defined by thresholds. Other patters are included too, such as leaking patterns (such as memory leakages) and emerging patterns.  

In conclusion, AIOpS is the infusion of AI into the cloud. This helps the platform become more adaptive and efficient. The Azure platform is always evolving and the system always needs to be improved to be maintained and up to date. Azure is not a small cloud ecosystem, we are talking about 160 datacenters that work on  60 regions. Currently, AIOpS is working hand by hand with the Microsoft Research team. They develop AI solutions for the cloud. AIOpS is based on data. The data is then transformed and distilled into insights. The insights are the most useful data for businesses, as it is what makes them advance.

DevOps
AI
Cloud
AIOpS
About the author
Lucas Bonder -Technical Writer
Lucas is an Entrepreneur, Web Developer, and Article Writer about Technology.

Related Articles