Our second day at Ignite was busy busy as usual! We’ve gotten into the full swing of things now with the demos, presentations, customer briefings and LEGO giveaways.
As expected, sessions we attended today were Azure focused, with an emphasis on proper (and hybrid) monitoring, cloud migrations, and new tech like Azure Arc.
BRK2208: Introducing Azure Arc
by Jeremy Winter, Microsoft’s Partner Director of Azure management, and Travis Wright, Product Manager of SQL Server
Considering this session was only announced yesterday, the full house turnout was impressive, even if somewhat expected given the buzz around Azure Arc. “We are taking a comprehensive, multi-dimensional approach with Hybrid,” said Winter when introducing Microsoft’s latest approach.
Azure Arc is a set of technologies that enables Azure data services and extends Azure management across your environment – working with Windows and Linux servers, Kubernetes clusters and data services running across datacenters, edge and multi-cloud.
In other words, it takes the Azure constructs down onto on-prem environments, which clearly embodies Microsoft’s new hybrid monitoring strategy. “Essentially it means you get cloud innovation,” says Winters. “It’s always up to date, it’s always been patched, you get a cloud billing model, and you can scale in seconds.” The data control plane is effectively extended so you can plug your own resources into it – like Windows Server, Virtual Machine, Linux – and then manage it in one place.
With many of us using the full spectrum of Azure services with (previously) no way of managing them all in a consistent way, Azure Arc is a welcome addition that lets us do exactly that across Azure, data centres, and out into the edge, taking the best practices from the cloud and applying them back into other environments in a uniform and consistent way.
After taking us through a couple demos and customer stories, Winters passed the mic on to Travis Wright who focused on what Azure Data Services brings to Azure Arc.
The first point he covered was that we can deploy anywhere. Using Azure Data Studio, we can deploy locally into your data centre regardless of where it is, and when that managed instance of SQL or Postgres is deployed into your environment, you get a similar service to what you get on Azure in the sense that it is constantly being updated. You get new features, bug fixes, security patches, elastic scalability, increased security and a unified management experience alongside a new cloud billing model.
He then proceeded to walk us through it all with a few demos.
OPS10: Building the foundation for modern Ops: Monitoring
by David Blank-Edelman, Microsoft
This was the first of a series of presentations that introduces the new Azure Admin to the world of monitoring. Called the OPS Learning Path, this introductory session introduces the fundamentals including Site Reliability Engineering, Service Level Indicators (SLI), Service Level Objectives (SLO), and some great advice to avoid the normal pitfalls.
Here’s what our Chief Product Officer John Shaw had to say about it:
“I went to a great session this morning on “Building the Foundation for Modern Ops: Monitoring”. David Blank-Edelman of Microsoft was not only a funny speaker but had put together a nice structure describing Monitoring as the foundation of an “operations pyramid” - kind of like Maslow’s hierarchy of needs – it all starts with monitoring, and until you monitor you can’t do anything else well. So true, but amazing how many IT and DevOps people either don’t monitor or don’t look at their monitoring data still.
He also showed quite a nice circle of “reliability factors” which were, if I remember rightly Availability, Latency, Throughput, Coverage, Correctness, Fidelity, Freshness, Durability. I am still trying to work out what Freshness exactly means …. The examples of how do use Azure Monitor were nice but all fairly introductory level given the time available – although I hadn’t seen before how you can generate some quite nice visualisations from the Resource Graph Explorer. Another great monitoring session and again well attended but lots of empty seats in a huge auditorium – the topic is starting to get hot but still not as much attention as it should get.”
For reference, here are a couple slides:
The OPS Learning Path (or as John puts it, an “ops pyramid”):
The different aspects of reliability. (Blank-Edelman noted that it's okay to spend most of your time in Availability monitoring!)
A little tip – always remember that:
Compared to Microsoft Ignite conferences from a couple years back (which were generally very lightly attended), this session’s high attendance rate is a clear indicator that Azure is gaining more-and-more credibility.
While the Azure Monitoring message from several years ago was very, well, confusing (to put it mildly), the message today plainly shows that proper monitoring (in the cloud or otherwise) is the future for today's IT Pro. As workloads move to the cloud, proper monitoring will be the only way to determine if something isn't working as expected.
THR2152: Azure tips and tricks: Become more productive with Microsoft Azure
by Microsoft Technical Product Manager Michael Crump and Microsoft Project Manager Vahe Minasyan
A demo-heavy session, Crump and Minasyan covered a range of tips and tricks for navigating the Azure Portal, making the most of the Azure CLI, and working with IDE and editors. Two tips stood out to us in particular.
The first is a new “interactive” feature in Azure CLI (currently in preview) that offers autocomplete and suggestions for commands. Achieve the same code in much quicker time! It’s also a great tool for new developers who are unfamiliar with Azure CLI.
The second covers using Azure automation and runbooks to handle some of the common day to day tasks like turning VMs on and off and deleting unused resources using tags.
Find the full collection of their 230 tips and tricks here https://microsoft.github.io/AzureTipsAndTricks/
THR3108: Azure Migrate: The next evolution
by Bharathram Sivaraman, Microsoft
This was an odd session for us, as what we were expecting to be a high level overview on how to migrate on-premise workloads to the cloud, was actually a session on an ISV tool that helps resellers migrate workloads using an Azure appliance. Don't get the wrong idea, this wasn't a vendor demonstration, but a session demonstrating a solution that vendors can use to help customers migrate their workloads.
Highlights included the following:
THR2266: The three things you wish you knew about, when moving to the cloud
by Erin Chapple, Microsoft
A Microsoft Mechanics Live session which was simulcast on YouTube (we haven't found the link yet, if one exists). Erin Chapple was very quick to keep to the Microsoft party line that the current cloud push is a "hybrid" push. In fact, her first tip (of two) was to not think of your resources as an island. E.g., Erin suggested that Admins should assess their landscape and use the Azure Network as your local network. The newly minted Azure Firewall functionality was highlight during the demo.
The second (and IMHO more important) tip was that IT Pros should not manage their cloud resources like their on-premises resources. Erin called out an Azure Migrate Assessment feature as being a great tool to assess your workloads before you place them in the cloud. Amusingly the Azure Migrate Assessment functionality was mentioned in the THR3108: Azure Migrate session, but either we missed it or the functionality was buried in the demo. Once the dust settles, it would be interesting to go back and review these announcements again.
There was supposed to be a third tip, but we might have missed it… 😊
Like yesterday, the booth was swamped!
As usual, we did some demos...
Some customer briefings...
A little interview with the Wall Street Journal...
Gave away lots of LEGO!!!
Was an amazing day all round! Come back tomorrow for more LEGO and good chats. See ya soon! :)