IT Certification and Training Blog

About Exam AZ-103: Microsoft Azure Administrator (Part 3)

Oct 15, 2019 3:23:54 PM / by Val Bakh

Read Part 1 and Part 2.

Part 3: Tips



This is the last article in a series about Microsoft exam AZ-103. Here we’ll provide some additional information that should help you to pass this exam.

Magic Wand

There is no magic wand, silver bullet, or other fairy-tale items that might work on Microsoft Azure exams. It is practically impossible to pass them without having the adequate knowledge. The labs are especially uncompromising. You can’t just click your way through the Azure portal without knowing exactly what you are doing. Even if you knew a lab task in advance, you may still get stuck in it if you are not sufficiently familiar with the way the portal works. For example, since the entire Azure world is built around the concept of virtualization, it is an easy guess that at least one of the tasks will involve the creation of a virtual machine (VM). If you only read about it but have not personally created at least one VM, you will almost certainly make one typical mistake. On the Management tab in the Create a virtual machine blade, there is a setting named Boot diagnostics. By default, it is set to On and therefore requires a suitable storage account to which the captured boot diagnostics are to be saved. However, no storage account is selected automatically, even if one already exists. When you click Review + create in order to initiate the creation of the VM, you will get an error message, saying that some required information is either missing or not valid. To avoid this error—or to correct it if it has happened—you should set Boot diagnostics to Off. This is a well-known issue, and you would definitely know about it if you tried to create VMs in the portal before. And if you didn’t, it might be next to impossible for you to figure out what’s wrong quickly enough.


New Features

Azure is an extremely dynamic environment. Whenever you sign in to the portal, there are almost always one or two notifications about the latest updates already waiting for you. The portal’s GUI keeps changing, and new features are added almost on a daily basis. Whatever materials you use for preparing for the exam—printed training manuals, practice exams, third-party blog posts, or even Microsoft’s original technical documentation on the Internet—there is always a good chance that some GUI-based procedures or information about some features at some point in time may be out of date.

Some things that were not possible in the past become supported. For example, managed disks as well as VMs and other resources that include managed disks could not be moved between resource groups in the past. However, starting from September 24, 2018, they can. Likewise, in the portal, some menu items are renamed or replaced and others are removed altogether. Options and settings are relabeled or moved around.

If on your exam you see a question that targets a recently added feature or that explicitly mentions some GUI elements, such as blades, pages, or settings, you should be very careful about how you answer. It is possible that the exam writers keep their contents up to date and that the question is based on the most current featureset or the portal GUI. It is equally possible that some questions about new features or related to the GUI remain static whereas the features and the GUI keep evolving. You should analyze the question, look for any tell-tales, and try to determine which state of the feature or version of the GUI the question is referring to.

In the labs, those issues are irrelevant because you are working in the real portal with the real, most current featureset and GUI. Therefore, get yourself a free-trial Azure subscription, if you still don’t have one, and practice all typical procedures regularly. Make sure you know exactly what is and what is not possible and where the most commonly used settings are located.



No one is perfect. Anyone can make mistakes, including technical writers—those who write Microsoft technical documentation and even those who write Microsoft certification exams. There are two main questions about the mistakes: (1) How do you know that something is a mistake? and (2) What should you do about a mistake when you discover one? The answer to the former is quite obvious: have a critical eye, question everything you read, and check out for yourself in the portal everything you realistically can. As for the latter, it is usually a bit more complicated, especially if it happens on an exam. Unfortunately, mistakes in exam questions are not entirely uncommon. So, what should you do if you get a technically inappropriate or even invalid question on your exam?

There are many ways in which an exam question can be wrong. None of the available choices might be correct. Or on the contrary, two or more choices might be equally correct instead of just one. A question might be ambiguous, and each of its possible interpretations might lead to a different answer. Some questions can even be outright technically invalid. Your most reasonable course of action is to analyze everything you possibly can about that question in those minute or two that you can spend on each exam question. Try to put yourself into the exam writer’s shoes, speculate about what the writer might have had in mind when writing that question. Sometimes, you might happen to recognize the scenario as one you had read about in some technical publication. You might have checked it out and found it was wrong, but the exam writer might have taken it at face value. If that is the case, select the answer that the article says is correct, even if it is actually not. In some cases, you might be able to figure out the writer’s train of thought and, hence, which answer the writer intended to be correct, even if in fact it was wrong. In the worst-case scenario, if nothing at all comes to your mind, just answer something. If you answer wrong, you won’t lose any points you have already scored. But if you “accidentally” pick the correct answer, you win.


“Minor” Things

In Azure, as in any other compute environment, there are many technical facts that are easy to overlook because at first glance they might appear too minor. Certain related resources must reside in the same region, whereas other related resources can be located anywhere. Some settings are inherited down the administrative hierarchy, whereas some settings aren’t. Some resources can be easily moved across geographic boundaries, some resources can be moved across only administrative boundaries, and some resources cannot be moved anywhere at all. To be able to navigate through your exam quickly enough, you need to pay close attention to such “minor” things when you study and should know them off the top of your head.

For example, once you have created a VM, you cannot rename it. You can change its computer name in the guest operating system or change its host name in DNS, but so far, there is no way to change the name of an existing VM without deleting it and creating another one instead from the original VM’s disks.

You can easily move a VM across subnets within the same VNet, but you can’t move it between VNets. Any VM created in accordance with the Azure Resource Manager (ARM) deployment model must have a network interface card (NIC). From the moment of its creation, any NIC must be connected to a VNet at all times, which means continuously, and it can be connected to only one VNet at a time. Hence you cannot disconnect a NIC from one VNet and then connect it to another one. If you could, the NIC would be left without being connected to any VNet during the switch. Consequently, when you create a VM, you must equip it with a NIC (which must be connected to a VNet), and once you have created the VM, you can’t move it to another VNet. As with renaming a VM, the only way to move a VM between VNets is to delete it and create another one instead on a different VNet.

You can create an empty availability set and then create new VMs as its members. But you can’t create a VM first and add it to an availability set afterward. A scale set is a similar type of resource, but its “rules of the game” are entirely different. For example, you cannot create an empty scale set and populate it later. All VMs in an availability set can be different, whereas a scale set can include only identical instances of the same VM.

There are hundreds or even thousands of such tiny pieces of information, and you need to know if not all of them, then at least the most common ones.



How do you know what is important for the exam and what is not? Strictly speaking, you can’t know that. However, there is a reasonably practical solution. Use Boson’s ExSim-Max product for exam AZ-103 as your primary preparation and practice tool. To take full advantage of this product, read the User Guide for it first and follow its recommendations. No one says it’s going to be easy, but with the right tools, techniques, and enough perseverance you can succeed.

Topics: Microsoft Exams, AZ103, Microsoft Azure Administrator, Azure, Microsoft Azure, Azure Labs

Val Bakh

Written by Val Bakh

Subscribe to Email Updates

Lists by Topic

see all

Posts by Topic

see all

Recent Posts