Boost Your AWS Infrastructure Performance — and Lower Your AWS Costs — by Detecting Low Network Utilization


nOps | , | February 29, 2020

When you set up an Amazon EC2 instance and attach an Amazon EBS volume to that instance, AWS will charge you for the instance and the Amazon EBS volume. And, if for any reason you have instances that are not actively churning business value, and you forget to stop the instances, you will continue to get charged. One of the methods to identify these unused EC2 instances is low network utilization.

Low network utilization is a symptom of unused Amazon EC2 instances that are still running. In an actively used Amazon EC2 instance, there should be good network I/O throughput showing both inbound and outbound traffic to that instance.

Low network utilization can also occur when an instance is running, but the network I/O throughput is quite low. If the instance isn’t running a workload that is critical to the business, it can either be snapshotted (for backup purposes) and terminated or stopped.

Identifying low network utilization

nOps cloud management platform has uniquely defined rules that can help you identify low network utilization. Then, you can take action to terminate the instance, or you can change the type of instance to reduce costs.

Here are the easy steps to identify low network utilization using nOps:

  1. Sign up to get an nOps account.
  2. Integrate your nOps account with your AWS account, so that you can obtain data for insights.
  3. On the main dashboard, click on Rules > nOps Rules.
  4. Click on the Cost tab to view information on your AWS Account.
  5. Under the list of resources found by nOps, you will find a line item that states the number of Amazon EC2 instances that have very low traffic.
  6. Click on that line item to see the resources and more details.
  7. You will then see a table that shows the Name, IP Address, and the instance types of the EC2 instances that have low network utilization. This table also shows you the estimated cost that can be saved from shutting down these resources.

Choosing the appropriate type of Amazon EC2 instance

When your Amazon EC2 instance is underutilized (the instance size is too large), you can change the type of instance accordingly. Amazon EC2 instance types are organized in instance families. Each family includes one or more instance sizes (to scale resources based on the needs of your target workload) and are categorized by their different capacities in terms of memory, CPU, and network I/O:

  • General-purpose instance family: Designed for generic workloads, this family is commonly used when deploying an application or service for the first time. While such instances are provisioned with a balance of CPU, Memory, and Disk I/O, they don’t contain the highest grade of CPU, memory, or other resources. The types of instances in this family are A1, T2, T3, T3a, M4, M5, M5a, M5n, and M6g. They are usually applied to web server workloads, containerized applications, data stores, cluster computing, gaming servers, caching fleets, and development environments.
  • Compute-optimized instance family: Used for compute-intensive workloads that require high-performance processors, such as high-performance web servers, scientific modeling, batch processing, distributed analytics, high-performance computing (HPC), machine/deep learning inference, ad serving, highly scalable multiplayer gaming, and video encoding. The types of instances in this family are C5, C5n, and C4.
  • Memory-optimized instance family: Used for memory-intensive workloads that process large data sets in memory and require fast performance, such as high-performance databases, distributed web scale in-memory caches, midsize in-memory databases, and real-time Big Data analytics. And, if you are going to run a Java application, high memory capacity will be needed to support the instantiation of objects in this Object-Oriented programming language (OOP). The types of instances in this family are R5, R5a, R5n, R4, X1e, X1, z1d, and High Memory.
  • Accelerated computing instance family: Designed to perform functions such as floating-point number calculations, graphics processing, or data pattern matching using hardware accelerators or co-processors instead of software running on CPUs. It’s used for machine/deep learning, high-performance computing, computational fluid dynamics, computational finance, seismic analysis, speech recognition, autonomous vehicles, and drug discovery. The types of instances in this family are P3, P2, Inf1, G4, G3, and F1.
  • Storage-optimized instance family: Designed for workloads that require high, sequential read and write access to vast data sets on local storage, with delivery of tens of thousands of low-latency, random I/O operations per second (IOPS) to applications. It’s used for NoSQL databases (e.g., Cassandra, MongoDB, Redis), in-memory databases (e.g., Aerospike), scale-out transactional databases, data warehousing, Elasticsearch, and analytics workloads. The types of instances in this family are I3, I3en, D2, and H1.

Frequently Asked Questions:

Q: What is Amazon EC2?

A: Amazon EC2 is Amazon Elastic Compute Cloud, a web service that provides secure, resizable compute capacity in the cloud.

Q: What happens when you terminate an Amazon EC2 instance?

A: EC2 instances are visible in the console for a short time after they’re terminated. When an instance terminates, the data on any instance store volumes associated with that instance is deleted. Deleted, terminated, or released resources can’t be recovered. You can’t connect to or start an instance after you’ve terminated it. However, you can launch additional instances using the same AMI.

Q: What happens when you stop an instance using the AWS CLI?

A: When an EC2 instance is stopped using the stop-instances command, the API request will send a button press event to the guest, which will stop various system services. Graceful shutdown is triggered by the ACPI shutdown button press event from the hypervisor. When the graceful shutdown process exits, the instance will shut down. Stopping a running instance is only supported by instances that were launched with an EBS-based AMI. An instance in the stopped state enables you to modify certain attributes of the instance, including the instance type.

Q: What is AMI?

A: An Amazon Machine Image is a special type of virtual appliance that to creates a virtual machine within the Amazon Elastic Compute Cloud. It is the basic unit of deployment for services delivered using EC2.

Q: What is an Amazon EBS Volume?

A: Amazon Elastic Block Store provides raw block-level storage that can be attached to Amazon EC2 instances and is used by Amazon Relational Database Service. Amazon EBS provides a range of options for storage performance and cost.

In conclusion

To boost your AWS infrastructure performance — and save on AWS costs — check for low network utilization. Use the nOps cloud management platform to detect low network utilization. Then, you can terminate the instance, or you can change the type of instance to reduce costs.

Want sharper visibility into low network utilization on AWS and see how much you can save? here to get started with a free 14-day trial of nOps (or click here to sign into nOps if you’re already a user) and review low network utilization in your account.