Episode 055: Orchestrating Containers on AWS

Mike Pfeiffer on December, 26, 2019

In this episode we catch up with Michael Wittig from from Cloudonaut.io about running containers on AWS, Infrastructure as Code, and much more.

Don’t forget to subscribe to our mailing list at cloudskills.io/subscribe for weekly updates, exclusive training, and advice on how to amplify your career.

Full Transcript:

Mike Pfeiffer:
All right and we’re recording, so I’m going to do a quick intro here. All right, everyone. Welcome back to another episode of Cloud Skills FM. Super excited to have you here as usual and on the episode today we are going to be talking about AWS. So I have Michael Wittig from cloudanaut.io and Michael, welcome to the show.

Michael Wittig:
Yeah. Thanks. Welcome.

Mike Pfeiffer:
Yeah. It’s great to have you on the show. We’ve been doing a lot of Azure episodes lately, so I’m excited to kind of switch gears and talk a little bit about AWS and for anybody that doesn’t know about your company, obviously the links will be in the show notes and all that kind of stuff, but maybe you could tell us maybe your background and what you do at cloudonaut.io.

Michael Wittig:
Yes. So I’m working together with my progress. We are a two person company, a small consultancy, focused on AWS, so we only do AWS all day. So we are doing a lot of project work for clients. We are also running a small SAS application, which is, it’s like chat bot that connects AWS with a Slack so if something goes wrong in your AWS account then we send you the information to Slack and then there’s some escalation logic and things like this going on.

Michael Wittig:
And besides that we are also writing a lot of blog posts. We are focused or we are aiming for two a week at the moment. So that’s one of the thing that we write it. Also, we have two books and a few video courses and things like this.

Michael Wittig:
So that’s kind of the work that we do and everything is related to AWS so that’s important, so we don’t have any experience with Azure, obviously, with Google cloud. Yeah.

Mike Pfeiffer:
Right. Got it. Yeah. And I’m seeing, I’m guessing you probably stay super busy with AWS because it’s just, you know, such a broad platform and so many people working on it so it’s cool that you’re putting out lots of content and working projects as well. What’s the new book about?

Michael Wittig:
Yeah. So our your book is … so the title is Rapid Docker on AWS and our idea was that we … we are pretty excited with the possibilities on AWS to run Docker containers. So there’s a service called Fargate, which basically provides the ability that you, you hand over an docker image to AWS and they run it for you so you can focus on the, on creating the image and as soon as the image is done, then AWS takes over.

Michael Wittig:
And in combination with a new database service they offer Aurora Serverless. You can connect to my SQL database. And also the database is completely managed by AWS and it also scales, depending on the load, it can shut down to zero, if it’s not used.

Michael Wittig:
So our idea was that okay, there are so many amazing features available on AWS, so it’s really now a good time for existing applications. So migrations to move to this kind of architecture because it’s a very low for the teams and it makes a lot of sense, I think, for existing applications to use Docker and Fargate and so you are independent from your cloud providers, but if you, for some reasons have to switch, you can around your Docker workloads on another platform, but it’s also a reasonably, or it’s within reasonable effort, you can migrate an existing applications inside a Docker container usually. So it also works for existing workloads.

Michael Wittig:
So that’s kind of, yeah, why we, the reason why we wrote the book and then our focus was to make it as easy as possible to get started for people that are not necessarily into AWS or Docker. So if your trust has an existing application and we have lots of examples for languages and frameworks, how to turn them into a Docker image or Docker images and then how to actually deploy them to AWS.

Michael Wittig:
And last but not least, we also add the pipeline so you can then for from this point on, you can kind of go into the continuous delivery mode and yeah, just benefit from all the features that we have today.

Michael Wittig:
Also, we have existing applications that a lot of teams have to run. Yeah.

Mike Pfeiffer:
Yeah. It sounds like a great book. I love that, you know, it’s kind of geared towards beginners because in my experience in the field lately, it’s definitely, you know, especially the past 12 months, past 18 months working with enterprises, it’s been really clear that they’re interested in containers but not everybody that’s been able to get into it. So it’s important to be able to ramp up and understand the basics and then understand how to leverage it in something like AWS. So that sounds like a good one for folks that check out.

Mike Pfeiffer:
And you mentioned Fargate. So when it comes to working with Fargate service and AWS to run your containers, maybe you could explain a little bit to the people listening how you would utilize that because it’s not kind of like the traditional way of doing containers on AWS, right? It’s a little bit different.

Michael Wittig:
Yeah. So previously, before Fargate, you at some point in time you always had to launch an EC2 instance virtual machine and this was your host and then you’re running your containers on this virtual machine. And they might want some orchestration layer on top of it, like ECS or if you like, you can also use a Kubernetes EKS service. But still, there are very EC2 instances involved.

Michael Wittig:
And then last year Fargate came around and now and the EC2 layer basically is removed from the customers so we don’t have to care about this anymore and you can just … basically, you ask AWS to really run your containers or not to run your EC2 instance and then run the container on top of it, but they execute the container for you.

Michael Wittig:
And the pricing is based on how many seconds you run the container so it’s as flexible as an EC2 instance in terms of the pricing model. But yeah, you don’t have to manage the operating system of the host and all those things. So it’s very convenient to use it actually.

Mike Pfeiffer:
Yeah. It makes a lot of sense because it seems like you don’t want to get bogged down with, you know, trying to spin up clusters of EC2 instances. Most people that are doing this are usually just, they want to focus on the application. It sounds like they can do that with Fargate. So if I heard you right there, you can do Fargate based containers with ECS as well as each EKS. Right. So whether it’s elastic container service, which is the proprietary orchestration service or Kubernetes as a service on AWS, right?

Michael Wittig:
Yeah. So that was that’s what announced, at least. So it doesn’t, so Fargate only works of ECS at the moment. It is announced that it will work with EKS and so there is [Reinvent 00:06:57] coming up next week so that’s the big AWS show where all the announcements are made. So we might hear an announcement next week that actually ship Fargate for EKS, but there was also an announcement for EKS that they are now so called managed instances where AWS takes care of the EC2 instances basically.

Michael Wittig:
So it’s not really Fargate, but even if you want to use Kubernetes, there’s also a possibility for you now since one week to have it fully managed by AWS, including the vertical nodes. So that’s also an option that is available on AWS.

Mike Pfeiffer:
Nice. Yeah, that’s interesting to hear kind of how they’re progressing with that. I was curious about that. That’s good to hear and I’m sure we’ll have tons of updates as the Reinvent conference takes off next week so it should be interesting to hear what happens. And when it comes to ECS kind of like, you know, do you think that there’s so much talk about Kubernetes, in your opinion, does it make sense for people to kind of focus on Kubernetes or is it, you know, good enough to just start with ECS or I’m sure that’s a common question, right?

Michael Wittig:
Yeah. So let me, Oh, yeah. So for most teams, I don’t really see why they should use Kubernetes. So if you really want to only run your containers on AWS then it is a much more efficient to do it with ECS and Fargate. If you really benefit from all the features of Kubernetes and also if your team is able to actually use them because they know about the technology, then it is also a good choice, I think.

Michael Wittig:
But most teams and most workloads that I see in my work with clients is that they just need a place to run the containers. They might not even really know what Docker is. So that’s where they are at the moment. And I mean, they don’t need Kubernetes. They first have to understand how to create a Docker image. And if this works then easily, then they are happy. So they don’t need all the features. They don’t need stateful applications like databases running on the container clusters because we have manage databases by AWS so we don’t want to do this on our own usually depending on like maybe the size of the company and things like this. But for smaller teams, I mean, why should you manage your own database on a Kubernetes cluster if you can get it as a managed service from AWS?

Michael Wittig:
So yeah, so that’s why for most clients, I don’t really see the benefits. I know that a lot of people talk about it and I also see a lot of people using it, but I don’t see a lot of people succeeding with it. So yeah. That’s interesting to see how this evolves. So.

Mike Pfeiffer:
Right. It is, yeah. I’ve been having this same conversation a lot the last couple months. You know, there’s a lot that goes into working with Kubernetes so it’s a tough hill to climb. It’s not for everybody. That’s for sure. And I think kind of the theme that we’ve been hearing a lot lately in discussions is that to your point, more people are going to use it as a managed service probably in the future just because not everybody can build a team that has the knowledge to be able to support Kubernetes at scale. So I love that that you brought that up and we’ve got easy ways to get started on AWS.

Mike Pfeiffer:
When you’re working with customers. I’m curious, what are you seeing people run into in terms of, you know, common things that are stumbling blocks, confusing points. Is there any patterns that you’re seeing in the field that you could share with us that might help somebody before they run into it themselves?

Michael Wittig:
Yeah. So maybe, so if you talk about the managed versions that are available on AWS, they are … so EKS is around for one year and so if you are used to how AWS works, they release a service and it is working for some use cases and then they basically improve the service that it covers more use cases over time. So if your use case fits into the limitations of the service at the moment, then that’s fine. And that also applies to the Fargate model.

Michael Wittig:
So for example, if your application requires a shared file system or something like this, then Fargate is not an option for you because it’s not supported. And so this is what we see enterprise clients running into. So they, I mean, a lot of applications rely on a shared file system there and that’s so, I mean, I ask does your applications rely on a shared file system? And if they say, “Yes,” then okay, we cannot go with the Fargate at the moment because it’s not supported.

Michael Wittig:
So I think those are the biggest limitations that we have. And then it gets better integrated into the AWS. So the Kubernetes, mismanaged Kubernetes stuff gets better integrated into the AWS ecosystem with the security groups, the IM permission, so that that has evolved a lot in last year.

Michael Wittig:
So I think at some point in time you can say that EKS will be integrated as good as ECS because ECS is fully integrated. If you run Fargate, it really feels like a virtual machine so you have all the security mechanisms in place that you had with virtual machines and that is also becoming true for EKS, but there’s still some open issues there.

Michael Wittig:
But yeah, I think if you wait a few more months, then all those things will be solved. So it’s a valid approach but also to use EKS on AWS. So that’s definitely true.

Mike Pfeiffer:
Right. Okay. Yeah. One of the things I was doing was just kind of checking out your guys’ websites before we jumped on this podcast recording. Tons of free content. So anybody out there that’s listening, you know, check the show notes because you’ll be able to go off and find Michael’s website. There’s just tons of awesome content.

Mike Pfeiffer:
And I have seen, you know, you guys have been putting out stuff on Pluralsight, video courses on different sites and stuff. One of the ones I noticed that you had out there, and I’ve seen it before, is the CloudFormation course in doing infrastructure as code. That’s stuff that our community spends a lot of time on. Could you talk a little bit about CloudFormation and maybe you know, share some opinions about the service and how you guys use it. I know you guys got a bunch of templates available for people as well.

Michael Wittig:
Yeah. So that’s also something I have to disclose. So I’m I not using Terraform. So I’m a CloudFormation user, as a 100% CloudFormation. So I know all the pain points and there are a lot of pain points, but in the last weeks we saw a lot of announcements from AWS then, which try to improve the situation.

Michael Wittig:
So first, CloudFormation is a great service. If you run or if you use features that are around for some time, then there will be no issues. So you will see great coverage of all the features of AWS. So if you, for example, see there was a new feature announced yesterday and then you might want to use it today, then the chances that it’s not supported in CloudFormation are very high at the moment. So it is … so if you follow the news with AWS and it’s sometimes a little bit frustrating because sometimes it takes really long time for the CloudFormation team to basically add the new features so that you can use it in CloudFormation.

Michael Wittig:
And the problem is that if you manage all your infrastructure with CloudFormation, then you basically have no other or no good choices besides waiting for them to implement it. And this changed in the last weeks, so it’s now easier to also provision resources outside of AWS and so that’s kind of … I think they are kind of going into the Terraform territory where you can also deploy other things that are not AWS related so this is now also possible with CloudFormation. The problem is the community has to evolve. So they are not yet so many in Terraform providers basically supported and probably we’ll see how this will evolve. Yeah.

Michael Wittig:
The other problem that we see when we kind of work with clients is that CloudFormation or Terraform, that’s not the problem usually. So the of problem is that with CloudFormation you have to, I mean, you have to understand the underlying AWS resources that you use because CloudFormation does not add any abstraction layers on top of the API basically, so you have to really know all the settings off your load balance or all the settings of the database and for example, spinning up a database is, I mean, there are 50 settings or something like this and you really have to understand how they work.

Michael Wittig:
So that’s the main challenge. Understand what you actually have to configure it because if you come from the graphical user interface and then a lot of stuff is hidden under the covers and you with just a single button click, you’ll sometimes deploy or create multiple resources and that’s not true. With CloudFormation there, you have to specify every single piece and that’s what we see as one of the main challenges. And that’s also why we provide the templates for CloudFormation in our repository. Because we also see that does a lot of repetition going on, so I mean, everyone needs a network, everyone needs a database, everyone needs elastic search clusters and things like this.

Michael Wittig:
And so there is not really a good place in AWS or there was no good place in AWS to kind of share those things and also AWS does not really seem to … so they provide us the building blocks. They don’t provide us the solutions. So we have to come up with the solution as customers.

Michael Wittig:
And yeah, so there are things going on in the AWS ecosystem as well. So they have something called CDK that was released, which also basically now adds abstractions so it gets easier to use it without understanding all the details. But it’s, again, a very, it was real or it was made general available a few months before so it’s a really new thing and not yet covers all the services. But yeah, there’s a lot of things going on in this space. Yeah.

Mike Pfeiffer:
Yeah. I’m always into the infrastructure code and really I’m really passionate and I spend a lot of my time in that area. I always geek out with it. So CDK, you mentioned, though the cloud developer kit, that is essentially using traditional programming languages to do infrastructure as code. Right? So that’s probably geared more towards developers would you say?

Michael Wittig:
Yeah. So I’m not 100% sure. So the main thing too to understand is that the CDK generates CloudFormation templates. So this means all the limitations I talked about still apply. So we still, so if this is not, if the feature is not supported in CloudFormation, CDK is not really helping us. I mean, they tried to do some tricks, but is not properly done if they do some tricks, because they, I mean, there are some ways to basically have something called a custom resource, but it’s a nightmare because you have to maintain it because you cannot easily switch to the officially supported resource type.

Michael Wittig:
But yeah, so that’s a problem and that problem still applies to the CDK. What the CDK gives us is a, depending on what program and language you prefer, they are able to generate and the CDK in multiple languages. So that’s from the technology perspective, the interesting point. So the CDK is available in multiple languages and they created some kind of layer that basically executes, I think, as far as I understand, no matter what programming language you use in the end it will execute Java script code for you. So they can easily add more languages and they will all behave in the same way. So, and what you get in the programming language of your choice is a SDL to create AWS resources.

Mike Pfeiffer:
Got you. Yeah. It makes sense.

Michael Wittig:
And so, that’s kind of the idea.

Mike Pfeiffer:
Right. Got it. Okay. Interesting. And then just to touch on something you mentioned a little bit ago, make sure I heard you right. And when you were talking about Terraform, is AWS investing more in Terraform now because their end users are more interested in it?

Michael Wittig:
So I’m not sure. So I know that, I think it’s some, you maybe correct me if I’m wrong, I think there are cloud providers who actively maintain their Terraform providers. I think Azure is doing a lot of work themselves. [crosstalk 00:19:26].

Mike Pfeiffer:
Yeah. The Azure guys are, I’m sorry to cut you off there, but the Azure team is doing lots with Terraform because they’ve had a lot of pushback from the community on … so the templating system in Azure right now is just JSON only. We don’t have the ammo like you also have in CloudFormation. And so, obviously, you know how that can drive people crazy. So there’s lots of people in the community on the Azure side that are doing lots of Terraform and Microsoft’s been really closely involved with Hachi corporate and trying to help that story. So I was just curious if AWS folks are kind of going down the same road.

Michael Wittig: Yeah. So as far as I understand, and also if, so I talked with, there’s a guy called Anton Babanko and he does a lot of the AWS provider work and he is not employed by AWS so he’s a also freelancer. And as far as I understand him, they AWS is not really involved in contributing to the Terraform provider. So they kind of prefer CloudFormation. But I know that there are lots of people using Terraform, so I don’t quite understand why they don’t help those people a little bit. But the funny thing or the fun fact is still Terraform is, or the Terraform provider usually supports new features within a day. And so they are faster than CloudFormation, which is maintained by professionals or by paid staff. So yeah, so that’s kind of interesting.

Mike Pfeiffer:
Oh, wow. That’s really interesting.

Michael Wittig:
Yeah.

Mike Pfeiffer:
That’s fascinating. Interesting. Okay. I was looking at your guys’ GitHub repository where you have all your cloud formation templates and I’ll link that up in the show notes, but it’s cool because not only are there a bunch of samples, but you also have like, you know, build status, you got your build badge. So obviously you’re, you know, doing some continuous integration, testing your templates. That was always a challenge when I was doing CloudFormation templates at a large scale. And one of the things that might be interesting for folks listening is that that’s a common question that pops up. It’s like how do I test my templates? And it looks like you’ve got Travis CI involved and you’re kind of going through that process. Any pointers to help people kind of navigate that process?

Michael Wittig:
Yeah. So what we do is, so the first two steps are we make sure that the, so we use [Yama 00:21:39] for the templates, so we make sure that it’s valid Yama and we also make sure that it applies to some style rules. So for example, we don’t want to have more or one tap is two spaces. In our case, we don’t want to have four spaces as taps. So things like this.

Michael Wittig:
And the second step is we use a tool called CloudFormation linter, so cfn lint. And that’s a project maintained by AWS and it makes sure that, or it checks if the CloudFormation is or looks valid without making a call to the API so it runs locally. So those are amazing tools and you can also integrate them into your development environment. So if you use something like vs code or sublime, you can you get like direct support or direct feedback in your template by writing if it’s a valid template and if there are some configuration errors for example.

Michael Wittig:
So the next step, and this is where it gets complicated and also expensive in terms of money and of time, at some point of time you have to spin up the template so you have to create a stack and see if it works.

Mike Pfeiffer:
Mm-hmm (affirmative).

Michael Wittig:
And we do this with some custom written Java code or we have also some templates where it’s a Java script that runs the tests. But what we do is basically whenever we merge into master we run all the tests and we spin up, I think it’s over 100 snacks at a time, and then we do some simple tests against them. So if it’s a, we have some SSH [inaudible 00:23:13], for example, then we see if we can make a SSH connection and if it’s a website VC, if it returns a proper statutes code and a proper answer and things like this.

Michael Wittig:
And this is really time consuming because for some resources, so for example, if you spin up a database cluster, so this can take 10 minutes. If you spin up a static website with CloudFront is takes 30 minutes.

Mike Pfeiffer:
Yeah.

Michael Wittig:
And so that’s … it’s not really something that, yeah, so we want to make sure before that with the linter and stuff that the templates are as good as possible and then only then we invest into in the tests and, and we burn thousands of dollars a month for this test. We also run them once a week to make sure that nothing breaks in between. We also found a lot of like timing inbox and things like this wherever things work sometime but not all the time and if some resources are not available. So it really helped a lot to improve the quality of the templates.

Michael Wittig:
But yeah, we still need run them in a single region. So it would be better to run them in multiple regions at the same time. But it then, I mean it, it just doubled your cost if you run in two regions and that’s, I mean, it’s an opensource project so we only random in US East at the moment.

Mike Pfeiffer:
Yeah, makes sense. That’s and everything you just described there, I’ve been through all that and it’s incredibly difficult to maintain templates that aren’t, you know, going to break a month after you write one or two weeks afterwards. So to your point, not just doing tests as you check your code in, but also after the template is spun up, went through all that and I can agree with you man, that it’s not easy to do that and I could see how you would easily spend a bunch of money, but it’s, you know, if a team is actually looking at doing infrastructure as code as in a serious way, they, they need to do it the way you’re doing it, right? Because to make sure that the templates are always going to consistently work every time you use it.

Michael Wittig:
Yeah. And that’s really also I think a space where there’s not a lot of good tooling available, so that’s …

Mike Pfeiffer:
Yeah. 100%.

Michael Wittig:
Yeah. I’ve also kind of … I was a little bit frustrated when I was looking for tools that could help me to do the tests and they’re not really that much tests available or frameworks that that can help us. So yeah.

Mike Pfeiffer:
Yeah. There was an open source project called task cat that was built by the quick start team and they open source that, I don’t know if you’ve looked at it, but anybody listening might want to check it out. I haven’t looked at it in a long time, but they basically had a kind of a task runner that would do something similar to what we’re talking about. It would spin up a template and go test the infrastructure I believe, or at least I would check the code, but that might be something to take a look at. I’ll put it in the show notes for folks listening.

Mike Pfeiffer:
But, you know, kind of taking another approach, switching gears just a little bit, but what do you think is a good resource, in addition to your website, for folks that are maybe getting started with this stuff, you know, containers doing more dev ops type of operations, infrastructure as code. What are some good resources we could start pointing people to?

Michael Wittig:
Yeah, so that’s a good question. So maybe what you could do if you are based in a city where there is a local AWS user group, then I would definitely recommend to check them out. So there are lots of user groups and you can, I mean, if you are familiar with other meet ups then you can just sign up on meetup.com mutually and then you can, I mean, if you go there three, maybe three times, then you will hear a container talk so you get an understanding about how it works.

Michael Wittig:
If you preferred to read then the AWS documentation is not bad. So it has a good coding getting started section usually. So that’s also something that you could check out. But it usually, I mean, they require some understanding of AWS. So for example, at some point in time they will make references to things like IM or security groups and if you don’t know all the concepts then it will be a hard way to figure out everything on your own.

Michael Wittig:
So if you’re not experienced with AWS at all, then I can recommend that you check out one of the example applications and also AWS, so if you check out the AWS labs organization on GitHub, they have lots of examples. For example … lots of examples, for example. So there are many examples and yeah, so I think this will be good places to start.

Michael Wittig:
And also if you want to spend some money, and of course there are all these video courses or there’s Pluralsight, there’s a cloud guru, and there are … how’s it called? Linux Academy. I mean, many places where you can watch videos that explain you how it works. So that’s also, I think , I really like to listen to videos for introduction and kind of sessions. And as soon as I have the overview then I like to dive into the documentation. Yeah.

Mike Pfeiffer:
Yeah. I’m the same way. I kind of start with that and then once I get past the documentation then I start going in and get the hands on practice and all that kind of stuff. So that’s a good, some good advice here. So along with your, I was just kind of looking at your book here, Rapid Docker on AWS. So that’s the new book that just came out. And along with the book, it looks like you’ve got a seminar as well, right?

Michael Wittig:
Yeah. So that’s, so that’s our new thing. So we, or we discovered it or we basically we discovered when we want to learn something new. So if I want to learn something new, I mean the problem is to sit down and take the time and read through all the stuff. And lots of times, you just, I mean, after one hour you just say, “Okay. This is too complicated. I will do it the way I always did it.”

Michael Wittig:
And with the seminar, our approach is that multiple people, so maybe 10 people will sign up for the seminar and they will all learn together so they will start at the same time and they all have the same goal. They want to understand how to run Docker on AWS.

Michael Wittig:
So we have a content that is released like every other day we release a new video and then you take your application to the next level so you will start with your existing application. If you don’t have an application, we provide an example application for and then on day one, for example, you will make sure that it is able to configure the application somehow using environment variables.

Michael Wittig:
Then next day, you will create the first Docker files for the Docker image. Then we will add Docker compose. And so step by step we will make sure that your application is turned into a dockerized application and then we able to move into an AWS.

Michael Wittig:
And if you have any issues in the meantime like while doing this, you can … so there’s a forum, a discussion board, where you can ask your questions. And then the idea is that the people help each other. So that’s usually how I really, if I really understand something or if I want to understand something, then it’s usually I create a talk for the people and try to explain it to them. And if you can explain it to someone, then you also have understand the topic.

Michael Wittig:
And so that’s our approach that the people will talk together and help each other. And then there’s also a weekly session where Andreas and I will have a video conference with all the people in the seminar and we will discuss the problems of the week and how the problems were solved and what approaches are there and what options you have and why this option is good. In your case, my … or not good. In this other case, for example.

Michael Wittig:
And so that’s kind of the idea that depending or based on the applications that the people bring with them to Docker, to dockerize on AWS, we can use those applications for the discussion. And that’s really interesting because then we have real world applications and this is a very good thing, I think, to learn a new topic.

Mike Pfeiffer:
Yeah, I agree. That structure of training is super powerful. The socialized learning is really big. We do a lot of the same kind of stuff here over in my company. And like you mentioned, you know, having a community around you of people that you could bounce questions off of is super important. The other thing is, you know, a lot of times it’s hard to absorb all the information in just one sitting like one week or watching one video so kind of stretching it out over a longer period of time seems to make sense for a lot of people.

Mike Pfeiffer:
But when it comes to certifications, it’s like lots of people are talking about AWS certifications. You see lots of people, you know, saying, Hey, I got all the certifications that AWS has. So I guess my question would be, are you kind of a fan of the certifications that AWS does? And if so, like what are good ones for people to keep an eye on?

Michael Wittig:
Yeah. So I’m certified myself. So I’m, how’s it called? Dev ops pro and solution architect professionals. I think those are the two certifications that I have and they require some other certifications to qualify for them. I think the idea of certifications is a good idea because it, I mean, you need some way to identify if someone knows about the topic or not and I think the certification is the easiest way to measure that.

Michael Wittig:
It is definitely possible to have a certification and to be, or if you compare two people with a certification, it doesn’t mean that they are equally skilled on AWS, but at least they have a certain understanding of how things work. So that’s definitely what a certification can do for us.

Michael Wittig:
And it also helps for, so I hear a lot of people doing a certification if they want to apply for a new job, for example, because those certifications are really … so the demand for certified people is really high because if you are … so if you look from the other perspective of the maybe the consulting company. So if you are a partner of the AWS partner program, AWS requires a certain headcount of people who have certifications and only if you all can provide this amount of people that you are qualified for the next level in the development, like in the partner program of AWS so that’s also a very big or an important reason for the companies to have people who have certifications, then to demonstrate to AWS that they are able to help their clients.

Michael Wittig:
So those are base, I think, the two perspectives. So one, it’s for the people who applies for the job but also for the company who looks for people so that’s why they prefer someone with a certification because they also need those certifications for AWS.

Mike Pfeiffer:
Got it. All right. That’s awesome. Well, Michael. Thank you so much for being on the show today. Like I said, we’ll link up all of the resources in the show notes. For everybody listening, go into your website, your new book, and I really enjoyed the conversation today. Michael, thanks so much.

Michael Wittig:
Yeah. Thank you very much. Bye.

Mike Pfeiffer:
All right. Perfect.

Weekly Updates Delivered to Your Inbox

Subscribe to CloudSkills Weekly and get access to exclusive training, updates on industry trends, and advice to amplify your cloud career.