Xebia Articles

Predicting incidents is the holy grail for modern IT organizations. It is no wonder that we see an increase in the usage of AI for prediction. An ounce of prevention is better than a pound of cure. But to what standard should we hold these predictions? When is a prediction important enough to wake you in the middle of the night? What is the level and type of information that ought to accompany a prediction? This blog introduces a new concept that is meant to raise the bar.

Big projects take forever, never get done, and lose relevancy for before anyone uses them. Building things nobody uses is a waste of time, and money, but we spend fortunes doing it, over and over again. There is a better way.

It's Thursday, 09:00AM. You arrive at work and see a huddle of people in the meeting room near your desk. Monitors are spread out across the room and engineers from the web team, Java team, .NET team and database team are staring intently at their screens. You forego your morning coffee for now and enter the room to see if you can help.

Root-causing and troubleshooting issues across the IT environment is a hard task as today’s infrastructure and applications span on-premise and cloud environments, and are more elastic, dynamic and distributed in nature. 

Think about the following questions before reading this article: (1) did you ever encounter an outage in a business chain where someone depending on the chain could not perform their job for hours? (2) Is there anyone who has insight in all dependencies between systems? (3) Is the status of each system available to everyone? (4) Do departments report a scheduled release to the depending departments? 

1. Don’t Know The Possible Impact of Deployments

Fast, short release cycles make it hard to keep track of your application and infrastructure landscape. A small change deep down in the infrastructure may seem innocuous, but can have a large impact over time. Today’s monitoring solutions should be able to provide insight which systems could be impacted before you deploy new software.

Ever heard of the English saying: "when it rains, it pours", meaning whenever something bad happens, it's usually not alone? It's an old saying, appearing in books written in the 17th century, but it applies equally today in the case of IT monitoring. 

There is a new trend going around in IT Operations, one which has been hard to miss: AIOps. Ever since Gartner coined the term, this approach of collecting IT Operations data and applying machine learning to it to gain insights to benefit your business has really taken off. While the benefits for Ops are fairly clear, will this trend also benefit developers? You bet! Read on to find out how.

If you think you don’t need Artificial Intelligence for IT Operations (AIOps) because you invested in a lot of IT Operations Analytics tools (ITOA), think again. AIOps and ITOA are not the same.

 Today’s applications and infrastructure are constantly evolving and continue to span multiple environments. As the complexity of the environments increases due to cloud and microservices initiatives, getting full stack visibility into these dynamic environments becomes more difficult.