Microsoft Azure Architect 70-534 Exam Experience and Tips

The information below covers my Microsoft Azure Architect 70-534 Exam experience. Following this I will post a list of my study materials, so keep checking back for updates!

One real positive for me when taking this exam was that I realised if you have an MCSA 2012, you do not need to take another Azure exam to achieve the MCSE title. Handy, especially as I have been pretty vocal about my thoughts on re-certification for versioned exams!

Microsoft Azure Architect 70-534 Exam Experience

Almost everything I read in the run up to taking the Azure Architect 70-534 exam, suggested that it was going to be pretty tricky. Many people suggested to me it was harder than typical MS exams. For those of us who are already a bit cloudy, harder than the AWS SA Associate exam but easier than the SA Pro.

My personal experience (having done both) was that it was a little harder than the AWS SA Pro exam, mainly in prep time and breadth of information, but and the reputation was perhaps a wee bit overblown. Don’t get me wrong, it was definitely tricky, but I have a sneaking suspicion that they may have dumbed it down a little in the past few months, as my experience did not quite match that of those who came before me!

tricksy Azure 70-534

The scoring methodology was WAY better than many other exams I have taken in the past (including from Microsoft). When you have a multi-part answer (e.g. choose 3 of 5, etc), then for each correct PART you get a point. In other exams, one wrong selection means “nil points”! In the 70-534 exam, I could have got one wrong selection in every multi-part answer, and still walked away with half or more of the points, which is AWESOME! This really took the pressure off!

The exam is very similarly formatted to most other MS exams, with a couple of notable exceptions. There is a section with standard multi-part, ordering, drag/drop, multi-choice as you would expect. Once this is completed (or perhaps before?), then you do a number of case studies. Note: Once you complete each case study, you cannot go back to it, however, the timing for the case studies was cumulative, so you don’t have to worry if one takes you a bit longer than another.

The number of questions I had in my exam left me with plenty of time, vs some of my colleagues who have done it in the past as well as since, who had 50% or more questions and case studies than me (I had 39 questions spread across all sections of the exam). I can only suggest that perhaps there have been some changes of late which mean you may or may not end up with more time per question.

It’s also worth noting that one or two of the questions I received were based on ASM (i.e. classic) instead of ARM! Not enough that it would be worth learning ASM, but don’t be surprised if something does come up.

legacy azure cloud asm classic mode

Exam Tips and Advice

Here are a few tried and tested tips for most exams as well as specific to the 70-534 exam (based on my experience):

  • Flip through the case study questions as you get to each one to get an idea of the kinds of questions being asked (e.g. security, authentication, networking, etc) so that you can bear these in mind as you read the case study.
  • Don’t worry too much about the clock, they give you plenty of time, especially as there is no specific time limit on the individual case studies (I think there may have been in the past?). For around the number of questions you are likely to get, this is loads of time.
  • Personal opinion: Old questions are dead to me! What I mean by that is that I don’t mark questions for review and once I click Next I never, ever, ever, ever, [ever!] go back. Chances are if I wasn’t sure about an answer and I go with my gut, it’s more likely to be right. If I sit there paralysed with indecision, I just waste time (or worse, potentially change a correct answer to an incorrect one!). By the time I hit the end of an exam I generally have a feeling whether I have passed or not, so going back to get a couple of extra points is a waste of time and I am just desperate to see the result! 🙂
    The one and only contradiction to this rule is if I come across a later question which immediately triggers me remembering something, or even blatantly answers a previous question by asking another. These are as rare as hen’s teeth though!
  • Finally, this may sound a bit cryptic, but I can’t go into any detail obviously due to NDA. All I can say is don’t get weirded out by what seems like an odd handful of questions at the start of the 70-534 exam. I got some which didn’t make sense to me at all until the end of the series (which doesn’t allow you to go back). I can’t go into more detail than that, but hopefully this preps you more than me, so you are not as surprised!
Architect Grumpiness

I do have one complaint about this exam which I will therapeutically air publicly now; why on earth as an “Architect” exam should anyone have to memorise the thousands of possible combinations of PowerShell commands, or indeed any commands whatsoever?! Fortunately, the percentage of the exam weighted towards this is small, but it is ridiculous IMO. 532/533, yes! 534? Stupid!

There also seems to be a key focus on understanding the exact specs of exact machine types. IMO this is also dumb as with any cloud platform you simply pull up your machine list and match the right machine at the time. Wasting time memorising the spec of every A-series, D-Series, etc machine is completely pointless, but is unfortunately required reading (at least as a minimum to remember the key “odd” ones, such as which provide RDMA).

powershelgl azure 70-534 exam tips

Anyway, all in all, a reasonably fair exam across a broad and relatively deep set of information and services. Best of luck to you, and if you found this article useful please leave a comment below! 🙂

Want to Learn More?

Part 2 of this article, my 70-534 exam study guide and all of my 70-534 study materials is available here:

Microsoft Azure Architect 70-534 Exam Experience and Tips

Azure, Certification, Microsoft , , , , , , ,