Audience & Preparation for the MCSE: Data Management and Analytics Certification

In a previous article, I covered some of my motivations for working towards the MCSE Data Management and Analytics certification. Here, I plan on going over how the certification is divided and how it may satisfy various personas: developers, administrators, and architects.

The prerequisite for the MCSE is the completion of the MCSA: Data Engineering with Azure. This certification is based on two exams:

1.  70-775 Perform Data Engineering on Microsoft HDInsight
2. 70-776 Perform Big Data Engineering on Microsoft Cloud Services

Microsoft divided the exams on the line of Open Source vs. Microsoft specific toolset. Both exams cover a wide array of technologies for storage & compute. Further, both exams cover respective batch and real-time technologies to enable engineers to develop data architectures following the Lambda architecture.

These exams are definitely applied in nature where you need hands-on experience in the various languages. Further, you require experience standing up the technologies on Azure using the various deployment methods (Powershell, ARM, .net, Portal, etc).

Once the MCSA requirements are met, you can write a 3rd exam from a series of electives to meet the MCSE requirements. I elected the to take 70-475 Designing and Implementing Big Data Analytics Solutions. This would wrap up the other two exams at a higher level where you have to make design decisions based on use cases.

Audience

Let's look at common data management roles and how each will do with each exam.

• Developer - Start with either 775 or 776 based on your background and the toolset you will be using. If you come from the open source camp, go with 775. Both exams are very broad and it will be difficult to learn all the technologies without prior experience or ability to apply on the job. Very few developers have exposure to batch AND real-time processing. In both exams, you will be stretched into doing administrator duties which I think is a good thing.
• Administrators - The cloud does not eliminate the need for administrators. There is plenty of content in all 3 exams that will be beneficial to admins. There is a strong focus on deploying and securing the various tools. Admins typically evolve away from the portal to use Powershell cmdlets and ARM templates. So you will need to become proficient in scripting & automation. The exams will stretch you to also learn how to transform data in various languages; for pure admins, this will be a steep curve.
• Architects - I think these exams are the easiest for architect/designer level roles. Let's face it, the technologies between the Open Source and Microsoft stack all build on top of a distributed processing foundation. The concepts are very similar with minor implementation details. The Microsoft PaaS offerings will have more limitations, but significantly reduce the overhead to manage. Again, covering all 3 exams will require a wide breadth of knowledge and hands-on experience.

Exam Preparation

First off, be prepared to do a lot of work to pass each exam because of their breadth. Microsoft had 2 choices when splitting the 775/776 exams:

1. Batch vs. Real-time
2. Open Source vs. Microsoft specific offerings

I think option 1 would have been easier for developers to prepare.

Anyhow, from a preparation perspective, you need to have prior experience with Hadoop, need to learn Azure, change how you think about data processing to deal with cloud peculiarities, handle hybrid scenarios (very few are all-in with the cloud, its usually an evolution), and be ready to wear multiple hats. No exam is specifically geared at a single persona (dev, admin, architect).

From an exam objective perspective, the first step is to go through EACH objective and map out documentation/tutorials to read and perform. The exams are broad and the coverage is actually as advertised in the exam outline.

Hands-on is critical to exams 775/776. You need to know the Powershell cmdlets and have used them. You also need to know a lot of syntax on various languages. Some of the exams are based on older versions of the toolset; so familiarize yourself in old and new (this applies primarily to ADF and some powershell).

Online Documentation

Microsoft's web site will be your greatest resource to learn the various technologies. The documentation is extremely comprehensive and well organized. The search function on Microsoft web site isn't great unless you search the outline....so revert to search engines instead. For most tools, you can export the entire documentation to PDF so you can read on the go which is a bonus.

For the exams involving HDInsight technology, I highly recommend you refer to Hortonworks documentation and the Hortonworks forum. Further, there is a TON of material online on Hadoop. Lastly, refer to respective Apache project wiki documentation.

Video Training

Beyond reading material, I leveraged online video from online training providers. I found the following providers to be excellent:

• Pluralsight - Has the most Azure and Big Data content by far
• OReilly / InfiniteSkills - Just like the books, OReilly rocks!
• Packt - Some are excellent, some are like their books....regurgitation of online documentation.

At this time, I was disappointed by the content from Wintellect / WintellectNow. Lynda had some good contents in some courses; I would not discount them.

Ebooks / Books

While videos can be useful, some prefer reading. Therefore, ebooks and books are a great alternative. Hadoop & Big Data have tons of books covering the subject while Azure is very sparse. The Azure books that I found were not very good (didn't add anything that wasn't on Microsoft web site). I recommend the following publishers:

• OReilly
• Apress
• Manning
• Wiley

I do not recommend Packt publishing books.

I have not reviewed the MS Press books, so I cannot provide a recommendation.

Happy learning! In the next articles, I will provide the specific resources I used for each exam, thus saving you time. Keep in mind that I had a lot of prior knowledge, so do not blindly follow my curriculum, do your own research and practice as well