You’re in the cloud! Now what?
Tips, tricks and tools to optimally run in the cloud.
If you’ve been following this migrating to the cloud series, you’re probably already (or nearly) running in the cloud, and that's great! But now what? The next step in the journey is about making sure your environment continues to operate as efficiently and cost-effectively as you planned for in the beginning.
In this article, we take you through a couple of tips, tricks and tools to help you ensure you’re running as lean as you possibly can, while not spending more than you have to – the key to optimally running in the cloud.
First things first, it’s important to review where you have room for re-platforming or re-architecting. This may take some effort to execute on, but as you know from our previous article, it is often the most rewarding option when it comes to managing operational and platform costs in the long run.
Jaco Venter, head of MServ, BBD’s cloud management team, comments: “An easy win we often see in deployments is to leverage Amazon Relational Database Service (RDS) as this helps reduce your operational overhead and could save you quite a bit of effort on an ongoing basis.”
Tip 1: SSM
On the note of saving time on operational overhead, another great tool to consider deploying in your environment is AWS Systems Manager (also known as SSM), which can be seen as the operations hub for your AWS environment. There are four key components to SSM:
- Operations management
- Application management
- Change management
- Node management
Venter says although all four areas are seen as valuable, he often sees clients benefit the most from the operations management, change management and node management modules.
Using these you can automate the maintenance of your servers, meaning you can centrally manage your backups and patching, again, making operations just a little bit easier. “This allows you and your team to focus on innovation and building new workloads that will help accelerate the growth of your business.”
Some of the other use cases for SSM include:
- Centralising operational data
- Automatically resolving application issues
Leverage operational data to easily manage applications and identify issues quickly across associated AWS resource groups.
- Implementing best practices
Automate reactive and proactive processes such as patching and resource changes to quickly diagnose and remediate operational issues before they affect users.
- Remediating security events
Adapt your security and compliance profile and analyse security events after the fact to prevent a future re-occurrence.
Here’s an overview of how SSM works:
Venter explains that SSM alleviates a lot of the operational overhead that teams often must do. The goals of using tooling, such as SSM in the case of AWS environments, is to make things easier for you and your team.
Tip 2: Rightsizing
From an operational perspective, rightsizing is an important process that BBD often encourages its clients to adopt, not just because it often results in a saved time, but often in a cost savings as well.
Rightsizing is the process of matching instance types and sizes to your workload performance and capacity requirements at the lowest possible cost. It's also the process of looking at deployed instances and identifying opportunities to eliminate or downsize without compromising capacity or other requirements.
“Once you’re in the cloud,” says Venter, “rightsizing is a key mechanism for optimising AWS costs, but it is often ignored by organisations when they first move to the cloud because they lift and shift their environments and expect to rightsize later.” He explains that speed and performance are often prioritised over cost, resulting in oversized instances and a lot of wasted spend on unused resources.
There are three areas that you should always keep an eye on:
- EC2 instance types;
- AWS storage classes; and
- RDS instance types.
“AWS has a tool called AWS Compute Optimizer, which gives downsizing recommendations within or across instance families, upsizing recommendations to remove performance bottlenecks, and recommendations for EC2 instances that are parts of an Auto Scaling group. It ultimately helps you identify where your environment could be running more optimally, and where you could be saving costs through rightsizing.”
When in the cloud, rightsizing is the most effective way to control cloud costs. It involves continually analysing instance performance and usage needs and patterns, and then turning off idle instances and rightsizing instances that are either over-provisioned or poorly matched to the workload. “Because your resource needs are always changing, rightsizing must become an ongoing process to continually achieve cost optimisation.”
Tip 3: RIs or SPs
After you have completed your rightsizing and know that the environment is running as efficiently as possible, it’s good practice to review how you can leverage either reserved instances (RIs) or savings plans (SPs) to further save money on an ongoing basis.
RIs have been around for quite some time, yet aren’t used as often as they should be. “I think it's a fear of the commitment that has to be made, I'm not sure, but the problem with this line of thinking is that you end up using the resources over the course of a year at an on-demand rate than if you’d committed to an RI – so why not commit to the year and get the benefit?”
Venter goes on to explain that RIs mean making a commitment to leveraging specific EC2 or RDS instances over either a one- or three-year period, and in exchange, AWS gives a discount on those instances. The best part is that these discounts can be up to 70%, depending on how you structure your RIs.
The only drawback Venter sees is that RIs can sometimes be cumbersome to manage and you don’t really get a tremendous amount of flexibility. To account for this, you can opt for convertible RIs as they give you some flexibility, but it can then become quite a big process when you want to convert them.
In 2018, AWS launched a new procurement vehicle called Savings Plans, not as an option to replace RIs but rather as an alternative. “Savings Plans provide a tremendous amount of flexibility,” says Venter. The key difference between the two is that with Savings Plans, there isn’t any commitment to a specific instance type or server; rather, it becomes a case of you committing to a specific amount that you will spend per hour on computing, and for that commitment, you receive a significant discount.
As always with your cloud environment, it's worth investigating which of these services make the most sense for your environment so that you’re able to reap the rewards and ensure a more optimally run cloud set-up.
Tip 4: Some words of wisdom from this series
- Consider your security goals and compliance requirements upfront;
- Understand what servers and applications you have running before you go down the migration route;
- Understand the six migration strategies and where they’re best suited;
- Ensure your operations are ready for move;
- Know your role to play in the shared responsibility model;
- Use tools to help minimise downtime; and
- Enlist the help of a partner experienced in migrating and optimising workloads in the cloud.
The biggest tip from this five-part series on migrating to the cloud is to enlist the guidance and assistance of a partner experienced in migrating and optimising workloads in the cloud. If you’re in need of such a partner – reach out to BBD.