EC2 instances not being correctly sized can quickly lead to unnecessary expenses. Rightsizing is critical to optimize costs and stability. 

However, rightsizing a single EC2 instance is one thing — everything gets much more complicated if you want to rightsize within Auto Scaling Groups. Most sources of rightsizing recommendations overlook this area — yet it is a huge portion of your compute cost.

Why is ASG rightsizing so hard to do right?

Rightsizing instances in an ASG is infinitely more complicated than rightsizing an EC2 instance. EC2 instances that are part of an ASG should be rightsized together — NOT individually as you would normally do when rightsizing. 

The dynamic nature of ASGs is such that instances come and go over time and may have different metrics distributions. Some terminated instances may have a higher or lower percentage utilization. These factors all add a huge amount of complexity to the calculations, particularly when it comes to Mixed-Instance ASGs.

To make reliable rightsizing recommendations, we need to account for (1) ALL of the instances that belonged to each ASG, both short-lived and long-lived, (2) track all of the instances’ metrics over time, and (3) group that data together to analyze their min and max resource consumption at an aggregate level.

Example of ASG metrics from the AWS console.

And if you make just one mistake and act on an unreliable recommendation, this may result in problems when the instances reappear — affecting the performance and stability of your workload.

nOps Makes ASG Rightsizing Simple and Seamless  

Tracking all of your instances, finding the right data, performing the right calculations, and accounting for all of the possible variables in a mixed-instance ASG is almost impossible to do manually.

That’s why nOps has integrated with the two industry-leading monitoring solutions, AWS CloudWatch and Datadog, for effortless rightsizing savings. We automatically analyze every EC2 instance in your environment (including shortlived ones) and pull their metadata to group them into their respective ASGs, analyzing min and max resource consumptions at an aggregate level to provide cost-saving recommendations.

Continuous coverage of resource-level insights such as memory, CPU, network bandwidth and storage are fed through nOps’s state-of-the-art ML engine for the best rightsizing recommendations available on the market. 

Rightsize with nOps for:

The most trustworthy rightsizing recommendations. Because nOps automatically collects and analyzes highly granular data, recommendations are 100% accurate and reliable — so engineers can act on them with the utmost confidence that workloads won’t be disturbed.

Up to 50% in immediate cost savings. When engineers don’t act on rightsizing recommendations, underutilized and idle resources continue to drive unnecessary AWS costs. nOps make it completely pain-free, safe and effortless for engineers to actually act on recommendations and start saving.

How it works

1. nOps integrates with your CloudWatch, CloudWatch Agent or Datadog to collect all of the metrics needed for ASG rightsizing recommendations, based on your last 10+ days of usage. Our API queries your data every 24 hours.
2. We quickly and efficiently process huge amounts of your CloudWatch data, crossed-referenced with AWS EC2 metadata and the latest AWS On-Demand pricing data to keep track of all of your ASGs (including terminated instances). These three sources are combined and fed through a Rightsizing Engine, allowing us to understand your dynamic ASGs holistically.

3. For each ASG, each of your instances is analyzed taking all relevant info into account, such as the metrics necessary for your particular operating system. For each instance in your environment, we make the following calculations:

   • Max Disk usage
   • Max Network usage
   • Max RAM utilization 
   • Max CPU utilization
     
     For each instance, our rightsizing algorithm compares maximum recorded usage against the capacity of a lower instance type, multiplied by a threshold value that accounts for potential future usage spikes. nOps takes into account the aggregate performance and utilization metrics of all instances within an ASG to make informed recommendations.
4. If all of the instances are rightsizable, the whole ASG is rightsizable. If you have several instance types, they can be analyzed and rightsized separately. 
5. These rightsizing recommendations are then pushed to nOps microservices, which are responsible for showing recommendations from the nOps platform on the UI.
   
   
   View your rightsizing savings in the nOps dashboard
6. Every 24 hours the process runs from top to bottom.

If you’re looking to consistently optimize your AWS costs, taking advantage of both Savings Plans and Spot is key — but balancing them can be complex. Not enough Spot means workload spikes result in expensive On-Demand coverage. And over-committing to Savings Plans can leave you paying for compute you don’t need. To make matters more complicated, some workloads are better suited for Spot and some are best for Savings Plans. 

nOps Compute Copilot offers proprietary ML-driven management and instance provisioning to put your workloads on the most reliable and stable compute at the best price in real time.

nOps intelligently balances your workloads between Savings Plans and Spot

AWS automatically applies Savings Plans to usage that has already occurred, prioritizing the highest discount rate. However, there are times in which you would actually prefer to push some of these workloads onto Spot (and out of the Savings Plan), so that the Savings Plan can be used for other resources that can’t be put onto Spot.

By proactively and strategically moving certain usage onto Spot, Copilot ensures that each workload is on the right type of discount to continually maximize your total savings. Copilot allows you to get discounts on:

• Harder-to-cover resources (for example, resources that can’t be put on Spot), so that you get discounts on ALL of your compute. 
• Resources outside of your connected clusters. Savings Plans apply across your organization. Copilot can drive certain usage to Spot, allowing resources even outside of your target workload to be covered by freed Savings Plans.

How it works:

While it’s very complicated to manually calculate how much of your Savings Plans to use to get a discount on all of your eligible compute usage, Copilot does it for you automatically. Let’s talk about how it works.

Compute Copilot ASG Lambda analyzes your AWS Savings Plans across your organization and your dynamic usage. Predictive ML is used to forecast your On-Demand usage and Savings Plans usage for the next hour to predict the amount of “Freeable” Compute Savings Plans.

If you have an unfulfilled Compute Savings Plans available and your ASG scales out with On-Demand instances, Compute Copilot Lambda will not move you onto Spot. 

If Copilot predicts there is some amount of Freeable Compute Savings for the next hour, it will automatically replace On-Demand with Spot when the On-Demand price is lower or equal to the predicted Freeable Compute Savings amount. As a result, it will free Savings Plans to cover some other previously uncovered On-Demand instance.

You can consult the documentation for more details on how nOps automatically and continually moves your workloads onto the most reliable and best priced Spot instances, balancing commitments and Spot for optimal price and stability.

Here at nOps we know the CUR like the back of our hand. We built a suite of enhanced features onto the CUR to help you transform millions of rows of contextless data into the who, what, when, and why of cloud spend. 

Enhanced Showback makes it easy to see where the money is. 

Easy-to-use and carefully curated filters help you to see clear trends and outliers in your financial reporting so you can allocate 100% of AWS cost to different workloads, environments, resource type, or other relevant categories.

Often, engineering teams trying to optimize for cost don’t have real-time visibility into hourly Savings Plan usage. It is very complicated to instantaneously calculate how much commitment is being consumed at any given time and the ideal amount of Spot to use — particularly across multiple AWS accounts.

The release includes new features enabling you to break down your costs by the most useful dimensions, such as:

• Hourly spend by purchase option for any of your environments. See at a glance how your AWS environment (or any slice of it) is balanced across pricing options (RI, SP, Spot…)

Hourly spend by Purchase Option

• The amount of your Savings Plans going underutilized during any given hour.

Amortized filters reveal the amount of Savings Plans you’re using (or not) each hour

• The tools (Databricks, Citrix…) provisioning compute in your environment. View hourly cloud spend by provisioner to pinpoint the products and activities driving your cloud spend. 

Break down cloud costs by tool or service

• The distribution of your compute spend across different AWS compute offerings.

Break down cloud costs by compute type

Check out the full blog for pro tips on how to use nOps to understand 100% of your AWS bill.

nOps was recently ranked #1 in G2’s cloud cost management category. Join our customers using nOps to understand your cloud costs and leverage automation with complete confidence by booking a demo today!

Many organizations are currently overspending on the cloud — and EC2 is typically the biggest culprit, driving 30-50+% of costs. 

To tackle this problem, Compute Copilot now integrates with ECS (Elastic Container Service), the leading native AWS EC2 management service. It automates the scheduling and scaling of your workloads to maximize stability and cost savings — all with minimal engineering intervention.

Compute Copilot makes it easy to confidently take advantage of Spot savings

Spot instances can save you 70-90% on your ECS costs, but using Spot manually with ECS alone is an extremely complex and time-consuming task. Here’s how Compute Copilot makes it easier:

Copilot navigates these challenges and more seamlessly on your behalf with automated real-time instance reconsideration. This proactive approach allows you to benefit from Spot savings effortlessly, with the highest standards of safety and reliability.

How it works:

1. Scaling Operation: AWS triggers a launch of a new instance due to the scaling operation by adjusting the DesiredCapacity of the ASG. This scaling operation could be by native AWS or from a custom controller
2. Compute Copilot Lambda Activation: Whenever the ASG (that is configured with Compute Copilot) launches a new On-Demand instance (e.g., in response to a desired capacity change), the Compute Copilot Lambda is activated.
3. Spot Instance Launch: The Compute Copilot Lambda responds by automatically launching a Spot instance with the mirror configuration of the On-Demand instance.
4. Attachment to ASG: The Spot instance is seamlessly attached to the ASG that is a capacity provider for the ECS cluster confirming its serviceability.
5. Graceful Task Draining and Instance Removal: With AWS Managed Task Draining enabled, the instance will be sent into a terminating:Wait state by AWS. This will drain all of the active tasks on the node.  

With Managed Task Draining, instances are gracefully terminated by (1) safely stopping running tasks, (2) launching replacement tasks on non-terminating instances, and (3) delaying instance termination until all tasks have stopped.

At nOps, our mission is to make it easy for engineers to optimize costs, so they can focus on building and innovating. With our platform, there’s no longer a reason to manually manage workloads; Copilot does it for you more effectively and at a lower cost. 

And there’s no vendor lock-in — Copilot updates configurations in your AWS-native tools,  meaning no major architecture update is needed to onboard or offboard. Plug it in or walk away at any time.

Join our customers using nOps to cut cloud costs and leverage automation with complete confidence by booking a demo today!

Compute Copilot makes it easy to save with Spot, bringing Spot market insights and awareness of your commitments to your ASGs. 

Copilot automatically and continuously tunes your configurations to ensure you’re (1) fully utilizing your commitments, and (2) always on the most cost-effective and reliable option available.

EC2 resource requirements are now automatically derived from prior ASG Launch Template configurations

Now, as Copilot moves your instances onto cost-effective and stable Spot, it will automatically ensure replacement instances are memory- and CPU-compliant. 

Simply select instance families for Copilot to use, and it will intelligently detect and apply minimum vCPU and RAM requirements from your prior ASG Launch Template configurations. As a result, there’s no longer any need to keep track of your workload requirements — Copilot fills in the correct parameters on the fly for seamless and reliable Spot savings.

If you prefer to exert granular control, you can also exercise the option to define exact parameters.

How it works:

When AWS ASG launches an On-Demand instance (e.g. m5.4xlarge):

• Lambda intercepts the instance launch event
• Determines the instance vCPU (16) and RAM (64 Gib) options
• Finds the cheapest stable instance type based on: instance families selected, vCPU and RAM options, nOps Spot recommendations
• Replaces On-Demand instance m5.4xlarge with the same or higher size Spot instance

For additional details, please consult the documentation. 

At nOps, our mission is to make it easy for engineers to optimize, so they can focus on building and innovation.

Copilot is built on your existing AWS-native ASGs. As a result, it’s ultra-easy to onboard and offboard. Simply plug it in to effortlessly run mission-critical workloads with peace of mind that you are scheduled on the most cost-optimized and stable option at all times. 

Join our customers using Copilot to cut cloud costs and leverage automation with complete confidence by booking a demo today!

As many tech organizations shift more resources to Kubernetes, cost-optimizing EKS is increasingly crucial. Yet, teams often lack the time and sophisticated tools needed to continually monitor and optimize cloud resources.

Recognizing this need, Compute Copilot now supports the two leading native AWS node management frameworks (Cluster Autoscaler and Karpenter) for a hands-off approach to workload management. 

Copilot automates the scheduling and scaling of your workloads for maximal stability, time savings, and cost savings — ensuring you are always on optimal cloud resource combinations with minimal engineering intervention.

Copilot makes it easy to save with Spot, while upholding the utmost standards of reliability.

Using Spot safely with Cluster Autoscaler can be a complex and time-consuming task, involving challenges like: 

• Spot interruptions disrupting workloads
• Continually shifting price and capacity in the Spot market
• The need to repeatedly balance Spot usage with existing Savings Plan and Reserved Instance commitments

Copilot navigates these challenges seamlessly on your behalf. Every 10 minutes, it analyzes the Spot market to predict termination 60 minutes in advance. It then automatically and continually moves your workloads onto diverse and less risky instance types, minimizing your risk of interruption. This proactive approach allows you to benefit from Spot savings, with enterprise-level SLAs for reliability. 

Copilot offers automated real-time instance reconsideration

Copilot continually monitors your dynamic usage and your existing SP and RI commitments, ensuring you are on the most cost-effective and stable blend of compute resources possible at all times.

At nOps, our mission is to make it easy for engineers to cost optimize, so they can focus on building and innovating. With the platform, there’s no longer a reason to manually manage workloads; Copilot does it for you more effectively and at a lower cost. 

And there’s no vendor lock-in — Copilot updates configurations in your AWS-native tools,  meaning no major architecture update is needed to onboard or offboard. Plug it in or walk away at any time.

1. Scaling Operation: Cluster Autoscaler triggers a scaling operation by adjusting the DesiredCapacity of the ASG.
2. Compute Copilot Lambda Activation: Whenever the ASG launches a new on-demand instance (e.g., in response to a desired capacity change), the Compute Copilot Lambda is activated.
3. Spot Instance Launch: The Compute Copilot Lambda responds by automatically launching a Spot instance configured to mirror the settings of the On-Demand instance.
4. Attachment to ASG: The Spot instance is seamlessly attached to the Cluster, confirming its serviceability.
5. Graceful Pod Removal: The Compute Copilot Lambda communicates with the Compute Copilot Agent running in the EKS Cluster to gracefully remove pods from the On-Demand instance before termination, preserving the integrity of your EKS environment.
6. On-Demand Instance Removal: The Compute Copilot Lambda terminates the corresponding On-Demand instance, completing the migration process.

For more on how to use Copilot and how it works, please consult the documentation. 

nOps was recently ranked #1 in G2’s cloud cost management category.

Join our customers using nOps to cut cloud costs and leverage automation with complete confidence by booking a demo today!

Rightsizing your EC2 instances plays a critical role in containing costs and fully harnessing the potential of your AWS resources. However, collecting the data needed to effectively rightsize can be difficult and time-consuming. 

nOps now integrates with your AWS-native CloudWatch, for effortless rightsizing savings. We now automatically analyze every EC2 instance in your environment for CloudWatch recommendations, which you can apply with one click on the platform.

For high-resolution recommendations, we also automatically ingest enhanced CloudWatch data from every instance with  CloudWatch Agents. Real-time coverage of resource-level insights such as memory, CPU, network bandwidth, volume size, and more are fed into nOps’s state-of-the-art ML engine — for the most reliable rightsizing recommendations.

Rightsizing with nOps for:

• The best rightsizing recommendations available. nOps automatically collects and analyzes highly granular data, for 100% accurate and reliable rightsizing recommendations.
• Significant time savings. nOps integrates with the two most popular AWS monitoring solutions (CloudWatch Agent or Datadog) and EventBridge. It automates the complex and time-consuming rightsizing process into a single click — freeing up engineers to build and innovate.
• Up to 50% in immediate cost savings. When engineers don’t act on rightsizing recommendations, underutilized and idle resources continue to drive unnecessary AWS costs. nOps make it completely pain-free and effortless for engineers to actually act on recommendations and start saving.

How it works

1.nOps integrates with your CloudWatch or CloudWatch Agent to collect all of the metrics needed for rightsizing recommendations, based on your last 15 days of usage. Our API queries your data every 24 hours and saves it in S3. 

CloudWatch metrics used for rightsizing

2. We quickly and efficiently process huge amounts of data from S3, cross-referenced with AWS EC2 metadata and the latest AWS On-Demand pricing data. These three sources of data are combined and fed through a Rightsizing Engine, to accept or reject each individual EC2 instance based on its utilization (resulting in recommendations for only underutilized resources). 

Each of your instances is analyzed taking all relevant info into account, such as the metrics necessary for your particular operating system. For each instance in your environment, we make the following calculations:

• Max Disk usage
• Max Network usage
• Max RAM utilization (depending on which version you have)
• Max CPU utilization

Our rightsizing algorithm compares maximum recorded usage against the capacity of a lower instance type, multiplied by a threshold value that accounts for potential future usage spikes. AWS Recommends using 80% of the lower size’s capacity as a benchmark. 

3. These rightsizing recommendations are then pushed to nOps microservices, which are responsible for showing recommendations from the nOps platform on the UI. 

The nOps dashboard shows your rightsizing savings

4. Every 24 hours the process runs from top to bottom. 

About nOps 

nOps was recently ranked #1 in G2’s cloud cost management category.

Join our customers using nOps to cut cloud costs and leverage automation with complete confidence by booking a demo today!

The Karpenter .33 Beta Release brings significant changes, notably transitioning from Provisioners and Node Templates to a more streamlined architecture of NodePools and NodeClasses. This change significantly alters the structure of Karpenter configuration files. For more on the changes and why they matter, check out our recent blog detailing all of the new Karpenter beta capabilities.

In this latest release, nOps Compute Copilot fully supports these advancements while ensuring backward compatibility with earlier Karpenter versions. Users can now effortlessly use Copilot regardless of your Karpenter version and upgrade at your convenience. Our backend handles all of the changes hassle-free on your behalf.

How it works:

1. We have upgraded the agent installed on client clusters. This agent is responsible for sending the CRUD commands to the Kubernetes API, choosing the proper CustomResources (NodeTemplates/NodeClasses or Provisioners/NodePools).
2. We have enhanced our API to validate the CustomResources based on the version. Therefore, we added a new set of validators for the NodeClasses and the NodePools. 

Compute Copilot now supports NodePools and NodeClasses

Whether your clusters operate on older or newer versions of Karpenter, or a combination of both, Compute Copilot is meticulously engineered to support you. 

Find out how Copilot makes it simple and easy to update to Karpenter .33 Beta in the documentation, which offers detailed information about the changes and upgrade procedures.

The latest update to Copilot for ASG features a new History of Actions Dashboard. It makes it easy to understand how Copilot is cost-optimizing your instances and how much money you are saving. 

You benefit from:

• Detailed workload tracking. View all the instances that were replaced by Copilot, and what they were replaced by.
• Cost impact analytics. Quickly understand the effectiveness of replacements and how long they ran.

Not yet using Compute Copilot to manage ASGs? It just takes a couple of quick tags to onboard and start realizing effortless ASG savings. 

If you’re currently using Spot to save on your ASG costs, you’re familiar with the complexity and time needed to cost-optimize your commitment usage and choosing the best Spot instances for your workloads. 

Compute Copilot was created to make it easy to save, bringing Spot market insights and awareness of your commitments to your Mixed-Instance ASGs. It automatically and continuously tunes your ASG configurations to ensure you’re (1) always using the right amount of Spot, and (2) always on the most cost-effective and reliable option available, gracefully replacing your nodes before termination.

And because it’s built on your existing AWS-native ASGs, it’s ultra-easy to onboard. Simply plug it in to effortlessly run mission-critical workloads with peace of mind that you are scheduled on the most cost-optimized and stable option at all times. 

New Mixed-Instance Policy & Lifecycle Hook Support

For hassle-free savings on the full range of your ASGs, Copilot now optimizes: 

• ASGs with Mixed Instance Policies — Launch multiple instance types and On-Demand Instances and Spot Instances within a single Auto Scaling group.
• ASGs with Lifecycle Hooks — Create solutions that are aware of events in the Auto Scaling instance lifecycle, and then perform a custom action on instances when the corresponding lifecycle event occurs.

It just takes a couple of quick tags to onboard and start realizing effortless Spot savings on all of your ASGs.

At nOps, our mission is to make it easy for engineers to optimize cost. Learn more about how Compute Copilot can help you save while freeing up your time to focus on building and innovation. 

How It Works

The below diagram provides additional information on how Copilot optimizes your ASGs.

ASG launches a new on-demand instance

Lambda intercepts the EC2 Instance State-change Notification event from EventBridge

If the created instance is not protected from termination and should be replaced, Compute Copilot performs the following steps:

1. Copy the Launch Template or Launch Configuration from the ASG launch template
2. In the copied Launch Template, modify Network Interfaces, Tags, Block Device and/or Mappings from the instance Launch Template or Configuration if needed
3. Fetch recommended instance types from nOps API
4. Request Spot Fleet with the copied Launch Template and recommended instance types
5. Once the Spot request is fulfilled, get the Spot Instance and wait for it to its state to be Running
6. Attach the created Spot Instance to the ASG
7. Wait for the attached Spot Instance’s state to be InService
8. If there is a lifecycle hook for termination for On-Demand, the On-Demand will go through the lifecycle hook action for termination
9. Terminate the On-Demand instance
10. If there is a Mixed-Instance Policy, modify the percentages of On-Demand and Spot accordingly.  

For more information, you can also consult the documentation.