AWS Solutions Architect Interview Questions (Master the Cloud Architect Interview Questions
Are you looking to start your cloud architect career? Do you desire to learn AWS solutions architect interview questions or cloud architect interview questions? If so, this blog is for you. Today we’re going to talk about some interview questions. The point is to help you build your technical competency. These questions are asked to gauge your levels of technical competency, communication skills, and ability to present complex technical materials.
Question 1. “What is the difference between IOPS and Throughput?”
This question is related to storage. If you’re going to build storage architectures, you must know, because different applications require different levels of latency versus different levels of throughput.
IOPS, or Input and Output Operations Per Second – determines how frequently you can read and write to the disc.
That’s the speed of disk access.
IOPS is related to latency, and is the amount of latency.
We can compare IOPS to the speed of a car that could drive at 200 miles an hour. The higher the IOPS: the more read/write operations per second, thus lower latency. NVME drives and SSD drives tend to have relatively low latency because the read/write operations are very fast.
Magnetic drives by comparison typically have much higher latency and much lower IOPS.
As mentioned, latency is measured in IOPS, and there’s an inverse relationship between the amount of IOPS and the actual latency on the network.
Throughput is related to the amount of data and is the amount of data that can be moved at any one period of time.
We can compare throughput to whatever the car’s trunk can carry. A car, a tractor-trailer, or a freight train would differ – you could carry a lot more stuff in the latter.
Video editors need drives that have very high throughput because they’re working with large file sizes. They can tolerate a little bit of latency.
A database needs extreme speed in terms of read and write operations per second, a higher IOPS but is not moving large amounts of data.
Question 2. “What’s the difference between a network ACL and a Security group?”
A network ACL, just like an access list on a router, keeps traffic outside of a subnet. Network ACLs are stateless! It means that nothing is tracking the state of the connection. Traffic that’s allowed in, is not being monitored, therefore you don’t know a lot of return traffic. That is why network ACLs need to be applied in both directions.
A security group keeps traffic out of a host, for example, a server. When you apply a security group to it, the traffic that you don’t permit will be denied. A security group is stateful! It means watching the state of the connection. It can see the data coming in here to a given host, is destined to the host and only the return traffic is allowed out. A security group needs to be enabled on the inbound direction.
Question 3. “What is the difference between RAID 0, RAID 1, RAID 5, and RAID 10?”
You need to know about the strengths and weaknesses of each approach because as an architect, you’re going to be dealing with this constantly.
RAID 0, or the most basic kind of RAID, is called striping.
If you have 3 discs, disc 1, disc 2, disc 3, data gets sent to disk 1, disc 2, disc 3 – disc 1, disc 2, disc 3 – disc 1, disc 2, disc 3.
The advantage is when you have 3 drives, of 2 TB each and a RAID 0 array, you have 6 TB, total.
The speed that you get is equivalent to 3 times each drive because you’re using all 3 drives in series. Write, write, write – Write, write, write – Write, write, write.
The disadvantage is It has zero redundancy. If 1 of the 3 drives in your RAID array fails, you lose everything because your data is spread across the drives.
RAID 0 gives you great speed and performance but has no redundancy.
RAID 1, is called mirroring.
If you have a 10 TB drive in your computer/server, and a second 10 TB drive, the data is copied from one drive to the other drive in real-time.
The advantage is that you always have an identical ready-to-use copy of your data. If one of the drives is lost, the other drive still has all the data.
The disadvantage is you don’t get any increase in capacity.
If you have two 10 TB drives, you only have 10 TBs of capacity, in total, because one drive is being always used for backup.
Also, you don’t have any speed improvement because everything is being written to one disc, and the speed limit is the disc, which is going to be written to the next disc at the same speed limitation on that disc, so you’ve got no speed performance.
RAID 1 gives you great redundancy and availability, but you do not get any speed or more capacity.
RAID 5, is the most common form of RAID in the enterprise environment, is called striping with parity.
AWS typically doesn’t recommend it on their network, but the entire enterprising world is running it. And I bet you, they’re probably running it on their internal RAID arrays for which they sell us EBS volumes as well as S3.
If you have 3 disks, disc 1, disc 2, disc 3, data gets written on all 3 discs, but they also send what’s called parity data (recovery) on all 3 desks.
Let’s say disc 1 gets data, disc 2 gets data, disc 3 gets parity.
The next time disc 1 gets parity, disco 2 gets data, disc 3 gets data. etc.
What happens is you’re taking one of the discs and you’re using it for recovery.
If there are 4 discs, you’ve got the capacity of 3, because 1 of them is going to be used for parity data. If you have 4 drives in a RAID 5 array, you’ll have 3 that’ll get used.
If there are 6 drives, 5 out of the 6 will be used and 1 will be used for parity.
The advantage is that RAID 5 generally speaking has some very good performance in terms of throughput.
It also provides great redundancy. If anything happens, basically you remove the bad drive, you pop a new drive in and you basically ask your RAID array to rebuild the data from the parity data from the other drives and you are good to go.
The disadvantage is that it can actually add some latency because writing this parity data definitely adds latency into the environment.
RAID 5 gives you a good blend of speed, performance, and redundancy.
RAID 10, combining mirroring and striping.
If you need more performance and lower latency than you could possibly get with RAID 5, there is another option.
The option is a combination of RAID 1 and RAID 0, that is RAID 10.
RAID 0 is super-fast because you’re running from drive to drive to drive.
RAID 1 is perfect for backup, have one drive here, it gets copied to another drive.
If you have 4 drives in the first RAID array and in RAID 0, you get 4 times the capacity and 4 times the speed.
If you made another RAID 0 array, you’d have again the same speed and capacity.
If you mirrored the first RAID array to the second RAID array, you’d effectively have 1 RAID array in terms of capacity and the other RAID array in terms of backup and redundancy.
The advantage is that it is a fantastic way for high performance.
The disadvantage is, that it requires double the number of disks and it gets very expensive very quickly.
RAID 10 gives you the speed of RAID 0, but with redundancy.
Question 4. “Can you describe the AWS shared responsibility model?”
What we’re looking for is to see if the interviewee actually understands what are the things that the organization would manage versus what are the things that AWS would manage?
Since the cloud is nothing more than a virtualized network and a data center. As an underlying technology, it’s just a network and a data center.
So as it stands, as a rule, AWS is going to manage the network and they’re going to manage all their data center things.
-The servers, the bare metal servers -they’re going to keep them patched with their BIOS updates, operating system updates, hypervisor updates. They’re going to take care of that.
-They’re going to make sure that their physical network, the kind that you’re riding underneath, is secure and locked down. They’re going to make sure that their network itself has good capacity, good availability, good redundancy, and can survive cable cuts and breaks and failures.
They’re going to take care of all that for you.
If it’s a serverless environment, they’re going to manage all the servers that are doing the serverless environment.
They’re going to secure the cloud and you’re going to secure your VPC and all your applications and services. That’s the shared security model.
AWS manages the data center, and you manage your stuff. But we like to ask that to see if people actually understand it because it’s very important.
The next two questions are actually very important questions!
They pop up in interviews. I’m going to include them here because lots of organizations, especially big organizations like AWS, are inclined to ask these kinds of questions.
Question 5. “Can you tell me about a major contribution you made to your last employer?”
We want to know, are you the kind of person that goes above and beyond? We have told you in the past what hiring managers desire – someone that goes above and beyond, communicates well, that’s energetic and enthusiastic, someone who likes to bring out the best in others, and is a problem solver.
Do you know who does this? People that make big contributions, because people that make these contributions make big contributions to their employer.
Tell them about a big project that you had and you took the lead on the project and you brought resources from this department or that department. Talk about how you met with the customer, and how you found the great customer requirements. Talk about how you designed something for the customer, with the help of this large team. And it delighted the customer so much, that not only did they purchase the solution, but they decided to use many more of your solutions in the future, or something to that effect. Show them something that you did that made a difference. That’s why – work hard, communicate with others, always try to bring out the best in others, develop your emotional intelligence, all these things matter.
Because when you can make a big contribution, your contributions will be known! Not only will you be hirable, but you’ll also be paid a lot more!
And what you did in one company directly carries over to another company, so make sure you’ve got a good list of major contributions you’ve made in your career!
Question 6. “Can you tell me about a major challenge at work? What did you do to get past it?”
There are lots of opportunities here!
-You can talk about a technology problem that the organization faced and an architecture that you designed to make it better.
-You could talk about a person that was on your team that maybe did not have the skills for the job and how you mentored them. And in the process of mentoring them, they were able to do their job and you made a great team, for example.
Give them an example of something that was really challenging, not your average ordinary things.
Show them what you did to defeat that challenge, how you rose to the occasion, solve the problem and how great it was for everyone.
See, employers want winners, as winners are people who had challenges and overcome them, moved on to the next great thing.
Cloud Architect Career Development Program
We’ll send you a nice letter once per week. No spam.