The cloud computing market is growing fast, and the trend is forecasted to continue. In fact, a recent Gartner report projected that global spending on public cloud services will grow 18% in 2021, to a total of over $300 billion. This means even more companies will accelerate their move to the cloud, and they’ll need employees with the know-how to make it happen.
The idea of helping a team reach a “baseline cloud fluency” is a phrase we hear a lot. You can have two engineers that excel at one kind of cloud operation, but if the rest of the team can’t speak their language, nothing gets done.
To better understand how we can help organizations get on the same page with cloud, we turned to our data. We analyzed 2.7 million responses to hundreds of questions across multiple areas of cloud expertise, homing in on the toughest questions, where the correct response rate fell below a C average (60%). From there, we identified key products, technologies, and topics that were often represented in those questions.
Informed by our research, here are four trip-up topics that we found to frequently stump cloud learners and what I think IT teams should know to master them, building on their cloud vocabulary in the process:
Auto-scaling helps to adjust your server count to manage shifts in traffic volume. For instance, imagine that you've just launched a new product, and an influx of users are excitedly trying to order it through your app or website. If you haven't configured auto-scaling correctly, you're about to collapse under the weight of your success.
Auto-scaling is tricky for a couple of reasons. Not only do you have to make the right call on which auto-scaling option you want to use, but you also have to ensure that you pick the right route. And it’s not easy -- our learners missed tough questions related to auto-scaling more than half (52.6%) of the time.
When you're configuring an auto-scaling group, ask yourself:
- How many servers do you want to maintain uptime?
- Do you want to adjust your server count manually?
- Do you want to schedule when to scale up or down, or,
- Would you like it to be based on conditions with your product performance?
If you think through these four parameters from the get-go, you can avoid problems down the line.
2. Identity and access management (IAM)
IAM allows you to manage who has access to a console by setting up users, groups, permissions, and roles. You can grant access to different parts of the platform, very granular permissions down to an individual user getting access to one service and not another. It’s how cloud resources speak to each other, how you audit them, and how you control access for your developers to update them, making it core to any cloud usage.
With IAM, the stakes are high: One compromised account could end up leading to a breach across your entire cloud footprint. This means getting IAM right has implications for everyone, not just the security team, yet according to our research, it’s a topic that tends to trip people up. Our learners missed tough questions related to IAM about half (50.5%) of the time.
To avoid an IAM blunder, you should always set up multi-factor authentication on root accounts and customize password rotations. Infrastructure as code, federated identities, and properly restrictive policies aren’t simple defaults to implement, but they will go a long way toward ensuring that your systems have exactly the access they need. Ultimately, there’s no substitute for a careful plan when it comes to laying out your IAM strategy.
3. Elastic load balancing
Elastic load balancing is designed to help you balance the network load across multiple servers. Our learners missed tough questions related to load balancers 53% of the time, and I think it’s difficult for two core reasons: First, you have to select the best load balancer; and then you have to enable features to make those load balancers more efficient.
When selecting a load balancer, consider which option is best suited to your applications and expected traffic. Next, you're going to have to address how you'd like to route your traffic to your various web services. Consider top-level configurations like sticky sessions, cross-zone load balancing, and path patterns. By addressing these configurations from the start, you will intelligently route traffic to optimize your performance for each web service.
4. Virtual private cloud
A virtual private cloud (VPC), also called a virtual private network, is like a virtual data center in the cloud. VPCs are the place to put your database, your application servers, your back-end reporting processes, and anything you don’t want directly exposed to anyone with an internet connection.
With VPC, your goal is to keep some information safe and off the internet while also connecting internet-facing services to both that information to the web simultaneously. You will likely want to set up your own VPC, versus relying on a preset, for improved security and customization. But it’s not easy; our learners missed tough questions 50.3% of the time related to VPCs. The whole process can get particularly hairy when you're configuring a custom VPC but don't have a ton of expertise in networking.
When setting up a VPC, there are many points where things can break down. It’s tricky. For instance, consider that there’s no transitive peering between VPCs on AWS. If VPC A can talk to VPC B, and B can talk to C, A still can’t speak to C. To master VPC, it’s all about recognizing these intricacies, exercising attention to detail, and practicing. That, in reality, is not unlike the process of learning any cloud skill.
The cloud landscape is vast -- and by tackling these terms you’re only scratching the surface. But, one skill at a time, you and your team can achieve baseline cloud fluency and maximize success in the cloud.
Ryan Kroonenburg is co-founder and instructor at A Cloud Guru: teaching the world to cloud.