See how the latest updates to Azure Migrate provide one unified environment to discover and assess your servers, databases, and apps, and easily migrate them to Azure. Jeremy Chapman is joined by Abhishek Hemrajani, Group PM Manager for Azure Migration, to explore how we’ve merged three different tool sets for discovery and assessment into one unified, at scale, agentless discovery and assessment experience.
Discovery: Track multiple SQL server instances on the same server, and discover the different user databases within those instances.
Assessment: Understand the compatibility of various applications running on the server for successful migration. Assess for Azure SQL, Azure VMware Solution or AVS.
Issues & warnings: Understand the readiness and any potential migration concerns.
Cost details: Review compute and storage costs for every instance that has been assessed. Estimate monthly compute and storage costs when you run your SQL server databases on Azure SQL.
Experiment & modify assessments: Modify your assessment properties at any time, and recalculate the assessment. Compare assessments across different dimensions that were used to create them.
02:01 — Discovery
04:32 — Assessment
06:28 — Issues & warnings
09:27 — Cost details
09:59 — Experiment & modify
10:47 — How to start
11:54 — Wrap up
For more on migration, go to https://aka.ms/LearnAzureMigration
We are Microsoft’s official video series for IT. You can watch and share valuable content and demos of current and upcoming tech from the people who build it at Microsoft.
- Up next on this special edition of Microsoft Mechanics we’re joined by Azure Migration and Business Continuity Engineering Lead, Abhishek Hemrajani to explore the latest updates to Azure Migrate to provide one unified environment to discover and assess your servers, databases, apps and migrate them to Azure. So Abhishek, welcome to Microsoft Mechanics.
- It’s great to be here, Jeremy, thank you for having me on your show.
- And thanks for joining us today. So about a year ago, you and the team actually set out to provide a centralized hub and unified tool set for migrating to Azure. And since then on Mechanics, we’ve actually shown how you can use Azure Migrate for everything from agentless discovery and assessment of your VMs and even bare metal physical servers, right through to right-sizing them and lifting and shifting them into Azure. We’ve also covered migration of VMware workloads, databases and apps, but what have you and the team been focused on lately with the latest round of updates?
- So we’ve been really focused on making it easy for you to migrate your VMs, your databases and your apps to Azure with Azure Migrate. Now with every release we set out to make these three primary migration scenarios as seamless as possible. Since discovery and assessment is the most important and critical phase of a migration project, we are now merging the three different tool sets for discovery and assessment into one unified, at-scale agentless discovery and assessment experience. First up in this experience is unified discovery and assessment for Windows server, Linux, and SQL Server. And soon we’ll expand capabilities to include .Net web apps that you could migrate over to Azure App Service.
- And that’s pretty significant because until recently you could do agentless discovery and assessment of your virtual machines and bare metal servers with Azure Migrate. But you would use integrations with things like Database Migration Assistant and App Service Migration Assistant to discover and assess your data in web apps. So, what have you done then to unify this experience? And can you walk us through an example of this?
- Absolutely Jeremy, let’s dive right into a demo and I’ll show you how things have gotten better. I’m going to start from the point at which you’ve set up the Azure Migrate appliance to discover your VMware VMs. These can be Windows or Linux VMs. Of course you can also discover bare metal servers with Azure Migrate. Now, if you provide us VM credentials we can inventory installed software and help you perform application dependency analysis without installing any agents. That allows you to discover network paths that originate and terminate on different VMs and the associated processes. Importantly, these credentials are never sent to Microsoft and remain on your Azure Migrate appliance running in your data center. With this release, we can now model servers running SQL Server and various instances and databases. Soon, we’ll do the same for IAS web servers. Azure Migrate supports both Windows and SQL authentication for discovering properties of SQL instances. Once the appliance is configured and discovery is initiated, Azure Migrate will be able to collect the required information to help you assess your SQL estate for migration to Azure SQL Managed Instance or Azure SQL Database. Here, you can see that we’ve discovered 21,364 servers. 980 are running SQL server. I can also see the OS distribution across my on-premise state. In the discovery view I can see servers which are running SQL server and how many instances are on each of the servers. So if you’re running multiple instances on the same server we’ll be able to capture that information.
- All right. So let me get this right. You can track multiple SQL server instances on the same server, but can you discover the different user databases within those instances?
- Yeah, absolutely. I can dive deeper into the instances to see the SQL version, the SQL edition, the count of the user databases and other SQL specific details. For example, here I can see that VM HRSQLVM04 is running five SQL server instances. I can see the various instances, SQL versions, count of user databases, SQL editions, and other details. We are adding support for SQL Server 2008 through SQL Server 2019. We will support the Enterprise, Standard, Developer, and Express editions. I can also go deeper into each instance and see the various user databases, their size, compatibility level, etc.
- Right. And previously, like you mentioned, you’d have to jump into quite a few different tools to achieve these types of results.
- That’s right. But to be clear, Data Migration Assistant as part of Azure Migrate is already a very powerful tool to assess your databases. What we are doing now is capturing the best of its capabilities to build a unified, at-scale discovery and assessment experience for SQL server. In addition to the discovery features there are also newer assessment types available for you. Now, we’ve always had Virtual Machine assessments, but you can now assess for Azure SQL, Azure VMware Solution or AVS, and very soon you’ll be able to perform application compatibility assessments for Windows Server. Now Jeremy, application compatibility is very important when you’re considering upgrading the operating system of your Windows Server, as part of the migration that helps you understand the compatibility of the various applications running on the server. Let’s create an assessment for Azure SQL. As I mentioned before, when you create an assessment for Azure SQL, we are assessing for both SQL Managed Instance and SQL Database. When you create an assessment most settings are applied automatically, but you can review and edit assessment properties. You can select the target location. In my case, I’ll stick with Southeast Asia. You can select the deployment type, which I’ll keep as the default recommended. I could choose Azure SQL Managed Instance or Azure SQL Database, if I want to go all in on one or the other. You can also select the options for reserved capacity, service tier and the performance history, which helps you better right-size your SQL workloads. I can then give this assessment a name and group the servers that need to be included. I’ll create a group and name it, ContosoVMSQL. Now I need to select my servers to assess. I will find all servers with SQLVM in their name and select all of those. Azure Migrate will ensure that servers running SQL instance are automatically included. Finally, I can review the details and create the assessment.
- Okay, so once you’ve run the assessment, what type of information do you get? And I remember with VM assessments you could get right-sizing, readiness, virtual machine sku and cost recommendations. Is that similar then for SQL servers?
- It absolutely is. For SQL, we’ll tell you the Azure SQL Managed Instance or Azure SQL Database Service tier that is the most optimal for your workloads. It takes only a few minutes for the assessment to get created. Once the assessment is created you can start understanding the Azure SQL readiness and Azure SQL cost details for your instances. You can see that in this case we assessed 57 servers running 207 SQL Server instances with 2,925 databases. You can see that the assessment gives you information about the readiness of your instances for SQL Managed Instance, SQL Database or where you may want to consider an Azure SQL VM. You’ll also get the aggregate compute and storage costs. So let’s go deeper into the Azure SQL readiness report. And you’ll see that for every instance you get readiness for SQL Managed Instance or SQL Database. We will also recommend which one might be a better target. And for that recommended target we’ll also tell you the optimal configuration.
- So how does Azure Migrate then determine these types of recommendations?
- That’s the magic Jeremy. Let’s look at the assessment for the instances for one of our Vms: HRSQLVM01. In this case, the VM is running three SQL instances. The first instance, MSSQLSERVER, is not ready for SQL Managed Instance. If you click on the recommendation detail we’ll tell you the rationale behind it. We classify migration information into issues and warnings. Issues are blocking conditions. For example, MSSQLSERVER has a database with multiple log files, which is not supported by Azure SQL Managed Instance. Warnings, on the other hand, are non-critical changes that may be needed to ensure that your applications run optimally post migration. In this case, you can see that there is a PowerShell job that you should review. Once I understand that Azure SQL DB is in fact the right deployment type for the instance, I can click on review Azure SQL DB configuration and get a consolidated view of all the databases along with the recommended SQL server tier for each of the user databases. Next, let’s look at the DMS2016SQL instance. This instance is ready for both SQL Managed Instance and SQL Database. I’m going to dive deeper into this instance to understand the reasoning behind the recommendation. The recommendation, in this case, is based on the configuration that will be the most cost effective. I can also see information related to total database size, CPU utilization, memory allocation and utilization, and IOPS. For this instance, there are no blocking issues but there are some warnings that I should be aware of.
- And that’s a lot of rich information and detail really to understand the readiness and any potential migration issues, but do you get similar details for cost?
- Absolutely Jeremy. Let’s look at the Azure SQL cost details. As you can see here, you can review compute and storage costs for every instance that has been assessed. This information allows you to estimate monthly compute and storage costs when you run your SQL server databases on Azure SQL. These cost components are calculated based on the assessment properties that you specify when creating the assessment, such as target Azure region, offers and important benefits such as the Hybrid Benefit.
- And it’s really great to see all the details and these assessments and the recommendations. But what if you want to maybe experiment and do different types of assessments, maybe you want to play around with what you might need in terms of compute or storage for specific workloads. Is there an easy way to kind of experiment and run different assessments against two groups?
- That’s a very important point Jeremy. You can modify your assessment properties at any time and recalculate the assessment. For example, here I’ll edit the properties and I can change the target deployment to Azure SQL MI and recalculate the assessment. Another useful feature is that once you’ve discovered your servers and have grouped them, there’s really no limit to the number of assessments that you can run on the same group. You can then compare the assessments across the different dimensions that were used to create them.
- Right. I can really see this information being super useful when you’re building out your business case. But what advice would you give people who are watching maybe starting out with their first discoveries and assessments?
- First start by prioritizing your discovery and assessment. Discovery and assessment is the most critical phase of a migration project, and will help you arrive at key decisions related to your migration. You also want to anticipate and start mitigating complexities. Think about architectural foundation and your management and governance approach in the cloud. And finally use an iterative approach to perform your migration. It’s all about creating the right size VMs that can help you execute more effectively.
- Cool stuff. Great progress since the last time we checked in on Azure Migrate but what are you and the team then working on next?
- So today we showed you unified discovery and assessment for SQL Server. Very soon, we’ll add capabilities for discovery, assessment and migration for .Net web apps over to Azure App Service with Azure Migrate. Also available in preview, is Azure Migrate App Containerization, which can help you containerize your .Net or Java web apps and deploy them over to Azure Kubernetes Service.
- Really looking forward to these capabilities. And thank you so much Abhishek for joining us today for the latest updates to Azure Migrate, but where could folks go who are watching learn more?
- So everything that they showed you today is available in preview. So you can try it out for yourself. You can also learn all things migration on aka.ms/LearnAzureMigration
- Thanks again Abhishek, and of course keep checking back to Microsoft Mechanics for the latest tech updates, subscribe if you haven’t yet and thanks for watching.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.