Menu

Solutions architect Neil Sayer talks about his experience attending this year’s Microsoft Ignite.

Day 1

This year’s Microsoft Ignite happened between the 26th and 27th of February, kicking off with a fast-paced introductory speech to prepare us all for two days of concentrated learning, and I think everyone left the arena feeling suitably fired-up.

My first session was “ARM Templates Tips and Tricks”, which was safe ground for me. This session provided some insight into what is often a bewildering subject for beginners, covering such advanced topics as user-defined functions and subscription-level deployments, having plenty to get your teeth into, even if you are an old-hand.

Microsoft has introduced the concept of “learning paths” this year, which allowed you to choose a set of presentations that are designed to take you through a technology or concept, from beginner to expert, over the course of the event. I had looked at these at the beginning but didn’t find any that appealed in their entirety, so opted to just pick and choose what interested me. I am glad I took this approach as people that stuck to paths reported a lot of repetition between sessions.

My second session was a swerve into modernising access and identity management with Azure Active Directory. This was an interesting lecture and concentrated on moving away from federation by using enterprise applications. This is something we have used quite a bit already, but the process is now more streamlined and easier to apply both within Azure and the SaaS applications themselves. They also demonstrated some of the “password-less” technologies, including using fingerprint on the Microsoft Authenticator app on phones to allow access to Office365. The end of this session was to show how much threat detection has been improved, which allows admin teams to spot suspicious activity far more easily.

Next, I went to another totally different session about using AI to augment data. This was a basic talk introducing the different forms AI can take, from ready-made models to making your own, and the different formats available from the different providers. It was a good introductory session and cleared up some misconceptions. It also showed that some of the high-concept ideas I had as to how we could use AI are not as simple as I thought.

My final presentation of the day was regarding serverless, using Azure Durable Functions, which was a good introduction to this form of serverless compute. The presenter worked for a company which creates software for evidence handling by the police.

Day 2

The second day began with Modern Operations, which focussed on developing a code pipeline in Azure DevOps, but also talked around DevOps as a practice –learning freely and failing together—, and the concept of Site Reliability Engineering. This was originally coined at Google but is getting wider acceptance as a development of DevOps practices and something I look forward to follow up on.

The penultimate session covered some concepts of monitoring, distinguishing between Service Level Agreements, Service Level Indicators and Service Level Objectives, and how these can be implemented in Azure. Also discussed were some future possibilities in monitoring, such as aggregating multiple services, correlation and anomaly detection, and monitoring the monitoring – how do you know when your monitoring has broken?

My final talk was much like the first lecture from that day in covering pipeline development in Azure DevOps, although it came at from a different approach, which covered the continuous improvement cycle. The session also explained why you might want to use DevOps Repos (good for internal projects) versus GitHub repos (good for projects with external collaborators).

This is just a flavour of the experience. Overall there was less in terms of new technology, but it gave plenty of food for thought and new approaches to previously known concepts. It was also interesting to see that this time there were sessions available on mental health issues in the workplace and other human-based problems rather than just technology.

About the author

Neil Sayer

Neil Sayer is a Solutions Architect who has worked at Eduserv in various roles for over ten years. During that time, he has worked closely with public and third sector customers to support and advise them in private cloud environments. In the last four years, when the opportunities for automation and infrastructure as code in public cloud became apparent, he became an enthusiastic convert and was involved in the product development of Eduserv’s public cloud offerings and its own internal, cloud-first approach.

To get our advisers working seamlessly alongside your own team, email us or call 0844 5000 115